Professional Documents
Culture Documents
Adin2111 2891062
Adin2111 2891062
Adin2111 2891062
ADIN2111
Low Complexity, 2-Port Ethernet Switch with Integrated 10BASE-T1L PHYs
Rev. 0
Information furnished by Analog Devices is believed to be accurate and reliable "as is". However, no responsibility is assumed by Analog
DOCUMENT FEEDBACK Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to
change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and
TECHNICAL SUPPORT registered trademarks are the property of their respective owners.
Data Sheet ADIN2111
TABLE OF CONTENTS
REVISION HISTORY
AVDD_H = AVDD_L = VDDIO = 3.3 V, DVDD_1P1 from internal low dropout (LDO) regulator (DVDD_1P1 = DLDO_1P1), and all specifications
at −40°C to +105°C, unless otherwise noted.
Table 1. General Specifications
Parameter Min Typ Max Unit Test Conditions/Comments
DIGITAL INPUTS/OUTPUTS Applies to the SPI pins, INT, RESET, and Px_LED_0 and
Px_LED_1 pins
VDDIO = 3.3 V
Input Low Voltage (VIL) 0.8 V
Input High Voltage (VIH) 2.0 V
Output Low Voltage (VOL) 0.4 V Output low current (IOL) (minimum) = 2 mA
Output High Voltage (VOH) 2.4 V Output high current (IOH) (minimum) = 2 mA
VDDIO = 2.5 V
VIL 0.7 V
VIH 1.7 V
VOL 0.4 V IOL (minimum) = 2 mA
VOH 2.0 V IOH (minimum) = 2 mA
VDDIO = 1.8 V
VIL 0.3 × VDDIO V
VIH 0.7 × VDDIO V
VOL 0.2 × VDDIO V IOL (minimum) = 2 mA
VOH 0.8 × VDDIO V IOH (minimum) = 2 mA
RESET Deglitch Time 0.3 0.5 1 µs
LED OUTPUT
Output Drive Current 8 mA VDDIO = 3.3 V
6 mA VDDIO = 2.5 V
4 mA VDDIO = 1.8 V
CLOCKS
External Crystal (XTAL) Requirements for external crystal used on XTAL_I/CLK_IN
pin and XTAL_O pin
Crystal Frequency 25 MHz
Crystal Frequency Tolerance −30 +30 ppm
Crystal Drive Level <200 µW
Crystal Equivalent Series Resistance (ESR) 60 Ω
XTAL_I, XTAL_O Input Capacitance (CIN,EQ) 1.5 pF Equivalent parallel differential input capacitance looking into
XTAL_I/CLK_IN pin and XTAL_O pin
Crystal Load Capacitance (CL)1 10 18 pF Including PCB trace capacitance and XTAL_I, XTAL_O
CIN,EQ
Start-Up Time 2 ms Crystal oscillator only
Clock Input (CLK_IN)
Clock Input Frequency 25 MHz Requirements for external clock applied to XTAL_I pin
Clock Input Voltage Range 0.8 2.5 V p-p AC-coupled sine or square wave at XTAL_I/CLK_IN pin
Clock Input Duty Cycle 45 55 % Input clock duty cycle 50%
XTAL_I Input Impedance (ZIN,EQ)
Driving Point Resistance (RP)2 6 kΩ RP||CP
Driving Point Capacitance (CP)2 3 pF
Jitter Tolerance (RMS) 40 ps
CLK25_REF Clock Output
CLK25_REF Frequency 25 MHz
VOH 1.05 V Load = 10 pF
VOL 0 V Load = 10 pF
CLK25_REF Duty Cycle 45 55 % Load = 10 pF
POWER-UP TIMING
Table 3. Power-Up Timing
Parameter Description Min Typ Max Unit
tRAMP Power supply ramp time 40 ms
t1 Minimum time interval to internal power good1 20 43 ms
t2 Hardware configuration latch time 6 8 14 μs
t3 Management interface (SPI) active 50 ms
1 The minimum time interval is referenced to the last supply to reach its rising threshold. There is no specific power supply sequencing required.
SPI
Table 4.
Parameter1, 2 Description Min Typ Max Unit
t1 SCLK cycle time 40 ns
t2 SCLK high time 17 ns
t3 SCLK low time 17 ns
t4 CS falling edge to SCLK rising edge setup time 17 ns
t5 Last SCLK rising edge to CS rising edge 17 ns
t6 CS high time 40 ns
t7 Data setup time 5 ns
t8 Data hold time 5 ns
t9 3 SCLK falling edge to SDO valid 12 ns
t10 3 CS rising edge to SDO tristate 15 ns
t11 3 CS falling edge to SDO valid (for readback MSB only) 12 ns
1 Guaranteed by design and characterization. Not production tested.
2 All input signals are specified with rise time (tR) = fall time (tF) = 5 ns (10% to 90% of VDDIO) and timed from a voltage level of 1.2 V.
3 Capacitive load on the SDO pin is 10 pF.
ESD CAUTION
ESD (electrostatic discharge) sensitive device. Charged devi-
ces and circuit boards can discharge without detection. Although
this product features patented or proprietary protection circuitry,
damage may occur on devices subjected to high energy ESD.
Therefore, proper ESD precautions should be taken to avoid
performance degradation or loss of functionality.
All tests were performed using a 1000 m 10BASE-T1L cable between Port 1 and Port 2 and DVDD rails powered from the internal LDO
regulators. Figure 5 only shows measurements for the 1.0 V p-p transmit level because AVDD_H = 3.3 V is required for the 2.4 V p-p transmit
level.
200
AVDD_H = VDDIO = 3.3V AVDD_L = 1.8V
190
DVDD1_1P1 = DLDO1_1P1
180 DVDD2_1P1 = DLDO2_1P1
170
POWER (mW)
160
150
140
130
120
TX = 1.0V p-p
110 TX = 2.4V p-p
100
–45 –29 –13 3 19 35 51 67 83 99 115
008
TEMPERATURE (ºC)
Figure 5. Power vs. Temperature, Single Supply Configuration, AVDD_H = Figure 8. Power vs. Temperature, Dual Supplies Configuration, AVDD_H =
AVDD_L = VDDIO = 1.8 V VDDIO = 3.3 V, AVDD_L = 1.8 V
250
AVDD_H = AVDD_L = VDDIO = 3.3V
240
DVDD1_1P1 = DLDO1_1P1
230 DVDD2_1P1 = DLDO2_1P1
220
POWER (mW)
210
200
190
180
170
TEMPERATURE (ºC)
Figure 6. Power vs. Temperature, Single Supply Configuration, AVDD_H = Figure 9. Power vs. Temperature, Dual Supplies Configuration, AVDD_H = 3.3
AVDD_L = VDDIO = 3.3 V V, AVDD_L = VDDIO = 1.8 V
250
AVDD_H = AVDD_L = VDDIO = 3.3V VDDIO = 1.8V
240
DVDD1_1P1 = DLDO1_1P1
230 DVDD2_1P1 = DLDO2_1P1
220
POWER (mW)
210
200
190
180
170
TEMPERATURE (ºC)
► 4 kB Port 2 PHY to host low priority receive FIFO Table 10. Port 1 and Port 2 Master/Slave Default Configuration
► 4 kB Port 1 to Port 2 transmit FIFO Port Master/Slave Configuration
► 4 kB Port 2 to Port 1 transmit FIFO 1 Prefer master
2 Prefer slave
TRANSMIT AMPLITUDE CONFIGURATION
AUTONEGOTIATION
The ADIN2111 supports two transmit amplitude modes of operation,
as follows: The ADIN2111 uses the autonegotiation capability in accordance
with IEEE 802.3 Clause 98, providing a mechanism for exchanging
► 1.0 V p-p and 2.4 V p-p mode (high level) information between the local device and link partners to agree to
► 1.0 V p-p only mode a common mode of operation. Single twisted pair autonegotiation
The high level transmit operating mode allows the ADIN2111 to is performed using differential Manchester encoding (DME) pages
support both voltage levels. The operating level can then be au- exchanged between the local device and its link partner. At a high
tomatically configured during autonegotiation (if enabled) based level, autonegotiation provides the following functions:
on the link partner capabilities. Note that in high level transmit ► Transmit
operating mode, AVDD_H must be supplied with 2.4 V or higher for ► Receive
the device to work properly.
► Half duplex
The mode of operation is configured through the P1_TX2P4_EN ► Arbitration
(Port 1) and P2_TX2P4_EN (Port 2) hardware configuration pin
signals (see the Transmit Amplitude section). The ADIN2111 also During the autonegotiation process, the local device advertises its
configures the default value for the transmit level register bits used own capabilities and compares them to those received from the link
for the autonegotiation process based on the level configured on partner. The arbitration mechanism defines the operating modes
those two pins (see the Transmit Amplitude Advertisement section). elected so that the transmit amplitude mode and master/slave
selection are configured for the linked devices.
The ADIN2111 is configured in high level transmit operating mode
by default on Port 1 (respectively Port 2) if the P1_TX2P4_EN If the link is dropped, the autonegotiation process restarts automat-
(respectively P2_TX2P4_EN) pin is left floating (internal pull-down ically. An autonegotiation restart can also be requested by writing
resistor). to the autonegotiation restart bit (AN_RESTART) in the BASE-T1
autonegotiation control register (AN_CONTROL) of the respective
MASTER/SLAVE CONFIGURATION PHY.
The 10BASE-T1L standard uses a master/slave clock scheme. This The autonegotiation process can take time to complete, depending
scheme is commonly used in full duplex transceiver standards with on the number of pages exchanged, but is always the fastest way
echo cancellation. to bring up a link. Clause 98 of the IEEE 802.3 standard details the
sequence timers and DME pages timing related to autonegotiation.
On a 10BASE-T1L link, one PHY is designated as the master, and
the other PHY as the slave. Autonegotiation is used to determine Autonegotiation is enabled by default for both ports of the
which PHY is the master and which is the slave. Master and slave ADIN2111, and it is strongly recommended to always keep it en-
assignment does not generally matter. abled.
► AN_ADV_B10L_TX_LVL_HI_REQ: advertisement of the the level to use on the PORT1 PHY (respectively Port 2 PHY)
10BASE-T1L high level transmit operating mode request bit based on the link partner capabilities.
► B10L_TX_LVL_HI: 10BASE-T1L transmit voltage amplitude con-
trol bit Determination of Transmit Level Resolution
Table 11. AN_ADV_B10L_TX_LVL_HI_ABL Settings For a 10BASE-T1L link, if either the local or remote PHY advertises
Bit Setting Description
that it is not capable of supporting the high level (2.4 V p-p) transmit
operating mode, or if neither the local nor remote PHY advertises a
0 Support 1.0 V p-p transmit level only request for high level (2.4 V p-p) transmit operating mode, the result
1 Support both 1.0 V p-p and 2.4 V p-p transmit level is operation at the 1.0 V p-p transmit level.
Table 12. AN_ADV_B10L_TX_LVL_HI_REQ Settings If both the local and remote PHYs advertise that they are capable
Bit Setting Description of transmitting in the high level (2.4 V p-p) transmit operating mode,
0 Request 1.0 V p-p transmit level
and if either the local or remote PHY advertises a request for high
level (2.4 V p-p) transmit operating mode, the result is operation at
1 Request 2.4 V p-p transmit level
the 2.4 V p-p transmit level.
High Voltage Transmit Level Request Thus, a PHY can ensure that the device must operate at the 1.0 V
Advertisement p-p transmit level. However, it can only request operation at the 2.4
V p-p transmit level.
The two PHYs of the ADIN2111 can be configured to advertise
a request for the 2.4 V p-p transmit level using the 10BASE- Software Configuration
T1L high level transmit operating mode ability advertisement bit
(AN_ADV_B10L_TX_LVL_HI_ABL) on the respective PHY. Note The ADIN2111 transmit level on the Port 1 PHY (respectively Port
that the 2.4 V p-p transmit level must be enabled using the 2 PHY) can also be configured in software using the following
P1_TX2P4_EN (respectively P2_TX2P4_EN) hardware configura- 10BASE-T1L autonegotiation advertisement bits on the respective
tion pin signal to enable the high voltage request advertisement on PHY:
Port 1 (respectively Port 2). ► AN_ADV_B10L_TX_LVL_HI_ABL: high level transmit operating
mode ability bit
Link Partner Transmit Level Advertisement
► AN_ADV_B10L_TX_LVL_HI_REQ: high level transmit operating
The high level transmit information advertised from the link partner mode request register bits
can be read using the following bits on each of the ADIN2111 The higher transmit level must be enabled on Port 1 (respectively
PHYs: Port 2) with the P1_TX2P4_EN (respectively P2_TX2P4_EN) hard-
► AN_LP_ADV_B10L_TX_LVL_HI_ABL: link partner 10BASE-T1L ware configuration pin signal to configure the two autonegotiation
high level transmit operating mode ability advertisement bits through software. See the Transmit Amplitude
► AN_LP_ADV_B10L_TX_LVL_HI_REQ: link partner 10BASE-T1L section for details.
high level transmit operating mode request If it is required to only operate the Port 1 PHY (respectively Port 2
These bits update during the autonegotiation process and are valid PHY) at the 1.0 V p-p transmit level operation, clear the respective
when the autonegotiation complete register bit (AN_COMPLETE) is AN_ADV_B10L_TX_LVL_HI_ABL bit, so that 2.4 V p-p transmit
set on the respective PHY. level operation is not advertised. In this case, autonegotiation can
only resolve to 1.0 V p-p transmit level operation, irrespective of the
Transmit Amplitude Resolution setting that the remote PHY advertises.
Those conditions can be set to enable an interrupt on the INT The ADIN2111 must be hardware reset to recover from a system
pin using the PHY subsystem interrupt mask register (PHY_SUB- error interrupt from one of the two PHYs (CRSM_IRQ_STATUS
SYS_IRQ_MASK). reserved bits read as 1 on the respective PHY).
Following a hardware interrupt on the INT pin, the interrupt source RESET OPERATIONS
can be checked using the PHY subsystem interrupt status register
(PHY_SUBSYS_IRQ_STATUS). The ADIN2111 supports the following chip resets:
► Power-on reset
Hardware Reset Interrupt ► Hardware reset
Each PHY of the ADIN2111 can also be configured to generate ► Software reset
a hardware interrupt after a hardware reset (RESET pin pulled ► MAC subsystem reset
low) by setting the CRSM_HRD_RST_IRQ_EN bit in the system ► PHY 1/PHY 2 subsystem reset
interrupt mask register (CRSM_IRQ_MASK) of the respective PHY.
All of these resets put the ADIN2111, including the two PHY cores
Although both PHYs can be used to generate a hardware interrupt, and the internal switch, into a known state.
it is recommended to use PHY 1 for this purpose.
Following a hardware interrupt received on the SPI host from the Power-On Reset
INT pin, the PHYINT bit (respectively P2_PHYINT bit) on the Status The ADIN2111 includes a power supply monitoring circuit to ensure
Register 0 (respectively Status Register 1) is also set to 1, notifying that the chip has the proper voltage supply before initiating the
an interrupt from PHY 1 (respectively PHY 2). power-up sequence. During power-up, the ADIN2111 is held in
The interrupt source can then be checked using the hardware reset until each of the supplies crosses its minimum rising
CRSM_HRD_RST_IRQ_LH bit in the system interrupt status reg- threshold value and the power is considered good.
ister (CRSM_IRQ_STATUS) of the respective PHY.
Hardware Reset
Software Requested Interrupt
A hardware reset is initiated by the power-on reset circuitry or
For system validation with an external host controller, each PHY of by asserting the RESET pin low for a minimum of 10 µs. The
the ADIN2111 can be requested to generate a hardware interrupt ADIN2111 includes deglitch circuitry on this pin to reject pulses
on the INT pin using the CRSM_SW_IRQ_REQ bit in the system shorter than 1 µs.
interrupt mask register (CRSM_IRQ_MASK). When the RESET pin is deasserted, all the input/output (I/O) pins
Although both PHYs can be used to generate a hardware interrupt, are held in tristate mode, the hardware configuration pins are
it is recommended to use PHY 1 for this purpose. latched, and the I/O pins are configured to their functional mode.
When all the external and internal supplies are valid and stable, the
Following a hardware interrupt received on the SPI host from the crystal oscillator circuit is enabled. After the crystal starts up and
INT pin, the PHYINT bit (respectively P2_PHYINT bit) in the Status stabilizes, the phase-locked loop (PLL) enables. After a delay of
Register 0 (respectively Status Register 1) is also set to 1, notifying 90 ms (maximum) from the deassertion of the RESET pin, all the
an interrupt from PHY 1 (respectively PHY 2). internal clocks are valid, the internal logic releases from reset, and
The interrupt source can then be checked using the all the internal SPI, PHY 1, and PHY 2 registers are accessible from
CRSM_SW_IRQ_LH bit in the system interrupt status register the SPI.
(CRSM_IRQ_STATUS) of the respective PHY. The CLK25_REF clock output stays low while the RESET pin is
asserted low and remains low for another 70 ms (maximum) after
PHY System Error Interrupts the RESET pin deasserts.
Each of the ADIN2111 PHYs can also generate system errors
interrupts. The interrupt flags are located within the reserved bit Software Reset
sections of system interrupt status register (CRSM_IRQ_STATUS) A chip software reset can be initiated by writing 1 to the SWRESET
of the respective PHY. field of the software reset register (SOFT_RST). Note that the
The system interrupt mask register (CRSM_IRQ_MASK) must be software reset does not reset PORT2 PHY. The PORT2 PHY can
configured on the respective PHY to allow system error interrupts. be reset using the hardware reset or PHY subsystem reset.
Refer to Table 212 for details on the interrupts mask. If a transmission is taking place when the SPI software reset is
initiated, the frame transmission stops abruptly and a runt or a
frame with a bad cyclic redundancy check CRC can be transmitted. The LED pins are suitable for ultra low power LEDs. The maximum
Once the chip is reset, the ADIN2111 is ready to bring up links. output current for each LED pin is 8 mA with a VDDIO = 3.3 V.
For higher LED power requirements, use an external transistor, as
When this software reset is initiated, a full initialization of the chip, described in the Transistor Controlled LED section.
almost equivalent to a hardware reset, is done. The I/O pins are
held in tristate mode, the hardware configuration pins are latched, The LED pins can also be connected to a host microcontroller
and the I/O pins are configured for their functional mode. The crys- general-purpose input/output (GPIO) (configured as a pulse-width
tal oscillator circuit is enabled, and after the crystal starts up and modulated input or hardware interrupt). This configuration is useful
stabilizes, the PLL is enabled. Approximately 10 ms (maximum) in applications where the user interface must be fully handled by
after writing the SOFT_RST keys, the internal logic releases from an external host controller (for example, an external LED module or
reset and the internal SPI, as well as the two PHY registers, are display).
accessible.
If the LED pins are directly connected to a host controller, place a
The system ready bit (CRSM_SYS_RDY) in the system status low value resistance in series between the ADIN2111 LED pins and
register (CRESM_STAT) on each PHY indicates that the start-up the host controller to avoid any potential current surge. The resistor
sequence is complete, and the respective PHY is ready for normal value must be defined based on host controller capabilities and the
operation. ADIN2111 LED pins output current capabilities listed in Table 1.
The CLK25_REF clock output remains low for 25 ms (maximum)
LED Pin Multiplexing
following a software reset.
For the Px_LED_1 pins only (P1_LED_1 and P2_LED_1), an inter-
PHY Subsystem Reset nal multiplexer must be configured to enable the LED signal on the
respective pin. Px_LED_1 signals are disabled by default and can
A PHY subsystem reset is initiated on the ADIN2111 PHY 1 be enabled using the DIGIO_LED1_PINMUX bits in the Pin Mux
(respectively PHY 2) by setting the respective PHY subsystem Configuration 1 register (DIGIO_PINMUX) within the respective
reset register bit (CRSM_PHY_SUBSYS_RST) to 1 in the PHY PHY.
subsystem reset register (CRSM_PHY_SUBSYS_RST). The reset
is applied for typically 1.2 µs, and then this bit self clears. All The Px_LED_0 pins (P1_LED_0 and P2_LED_1) do not need
of the PHY 1 (respectively PHY 2) digital circuitry is reset, and multiplexing.
any available active link drops. The PHY 1 (respectively PHY 2)
subsystem reset does not alter the values of the management LED Polarity
registers, which remain accessible throughout the sequence. The
subsystem reset is a short reset and can be used to put the device The four LED pins can be configured to support various LED circuit
into a known state while retaining the internal register contents. polarities using the LED polarity register (LED_POLARITY) on the
respective PHY. Three polarity modes are available for each LED,
MAC Subsystem Reset as follows:
A MAC only software reset can be initiated by writing the required ► Autosense (default)
pair of keys to the software reset register (SOFT_RST). The reset ► Active high
is applied for about 1.2 µs. The MAC subsystem reset interrupts ► Active low
any transmit/receive packet exchange between the MAC and the
In autosense mode, the ADIN2111 automatically senses the pin at
two PHYs, but does not drop any existing link or prevent a link from
power-up or reset to select the appropriate polarity configuration.
being established. The MAC subsystem reset does not alter the
values of the PHY registers. In active high mode, the ADIN2111 is configured to drive the LED
from the anode side.
Note that PHY 1 must be out of software power-down for the MAC
subsystem reset to take effect. In active low mode, the ADIN2111 is configured to drive the LED
from the cathode side.
STATUS LEDS
Example circuits are described in the LED Circuit Examples sec-
The ADIN2111 provides two configurable LED pins for each port: tion.
P1_LED_0, P1_LED_1, P2_LED_0, and P2_LED_1 (also noted as
Px_LED_x).
LED Function
The LED pin signals can be used to connect external LEDs to
The LED pins can be configured to display various activities of
indicate the ADIN2111 link status and transmit or receive activity.
the ADIN2111 using the LED function feature. The LED function
The activity assigned to each LED is configurable through the LED
control register (LED_CNTRL).
is configurable for each LED using the LED0_FUNCTION and ► LED Mode 1: blink duty cycle defined using the LED 0 on/off
LED1_FUNCTION bits in the LED control register (LED_CNTRL). blink time register (LED0_BLINK_TIME_CNTRL)
► LED Mode 2: blink duty cycle automatically defined by the
Note that the 7, 8, 9, and 10 (decimal) bit settings for LED0_FUNC-
TION and LED1_FUNCTION are not available in LED Mode 2. ADIN2111 based on activity level (%)
For each of the ADIN2111 LED pins, the activity behavior can be See Table 15 for the configuration options of the Px_LED_x pins.
configured using the two LED modes, as follows:
Table 15. LED Pins Configuration Summary
Parameter Px_LED_0 Px_LED_1
Pin Number 19 (P1_LED_0), 47 (P2_LED_0) 21 (P1_LED_1), 48 (P2_LED_1)
Internal Pull-Up or Pull-Down Resistor Pull-up Pull-down
Status at Power-Up or Reset Enabled Disabled (via pin mux)
LED Pin Mux Not applicable DIGIO_LED1_PINMUX bits in the Pin Mux Configuration 1
register
Enable LED LED0_EN bit in the LED control register (default: enabled) LED1_EN bit in the LED control register (default: LED_1 output
not enabled)
LED Polarity LED0_POLARITY bits in the LED polarity register (default: LED1_POLARITY bits in the LED polarity register (default:
autosense) autosense)
LED Mode LED0_MODE bit in the LED control register (default: LED Mode LED1_MODE bit in the LED control register (default: LED Mode
1) 1)
LED Function1 LED0_FUNCTION bits in the LED control register (default: LED1_FUNCTION bits in the LED control register (default:
LINKUP_TXRX_ACTIVITY) TXRX_ACTIVITY)
LED Blink Rate LED0_BLINK_TIME_CNTRL in the LED 0 on/off blink time LED1_BLINK_TIME_CNTRL in the LED 1 on/off blink time
register register
Maximum Current2 8 mA at 3.3 V 8 mA at 3.3 V
1 The 7, 8, 9, and 10 (decimal) settings for the LEDx_FUNCTION bits are not available in Mode 2.
2 See Table 1 for details.
POWER-DOWN MODES power state, and the PLL is active and can provide output clocks if
configured to do so. Any signals exposed to the MDI pins, P1_TXP
The ADIN2111 supports the following two power-down modes. and P1_TXN (respectively P2_TXP, P2_TXN), are ignored and any
► Hardware power-down active link on the respective PHY is dropped. The SPI registers are
► Software power-down
accessible, and the device can be configured using software.
Hardware power-down mode achieves the lowest power consump- The PHY 1 (respectively PHY 2) can be configured to automati-
tion. In this mode, the ADIN2111 is fully turned off and the SPI and cally enter software power-down mode after power-up, hardware
the two PHY registers are not accessible. reset, or software reset using the P1_SWPD_EN (respectively
P2_SWPD_EN) hardware configuration pin signal. The ADIN2111
can also be instructed to enter software power-down mode by set-
Hardware Power-Down Mode
ting the software power-down bit (CRSM_SFT_PD) in the software
The hardware power-down mode can be used when no operation power-down control register (CRSM_SFT_PD_CNTRL) within the
is required on the ADIN2111 and the power consumption needs to respective PHY.
be minimized. The device enters hardware power-down mode when
The software power-down status bit (CRSM_SFT_PD_RDY) in the
the RESET pin is asserted and held low. In this mode, all analog
system status register (CRSM_STAT) indicates that the respective
and digital circuits are disabled, the clocks are gated off, and all
PHY is in software power-down mode.
the I/O pins are held in tristate mode. In this mode, the ADIN2111
power consumption is equivalent to the internal circuit leakage. The The ADIN2111 PHY 1 (respectively PHY 2) exits software power-
internal registers are not accessible in this mode. down mode when the CRSM_SFT_PD bit is cleared within the
respective PHY. After exiting software power-down, and if autone-
Software Power-Down Mode gotiation is completed, the PHY 1 (respectively PHY 2) attempts to
bring a link up.
Software power-down mode can be used on each PHY to configure
the ADIN2111 registers before bringing a link up. In this mode, the Note that it is necessary to bring PHY 1 out of software power-down
PHY 1 (respectively PHY 2) analog and digital circuits are in a low mode to establish a link on PHY 2.
The ADIN2111 can operate in unmanaged or managed configura- ► PHY 1 transmit amplitude configuration
tions with the use of the hardware configuration pins. ► PHY 2 transmit amplitude configuration
The hardware configuration pins are standard pins with an alternate ► OPEN Alliance or generic SPI
bootstrap function. The ADIN2111 reads the configuration pin level ► SPI CRC/protection
immediately after power-up, hardware reset, or software reset,
and configures the two PHY settings accordingly. When active, All of the hardware configuration pins have internal pull-up or
the ADIN2111 immediately attempts to bring up a link on each pull-down resistors, as shown in Table 8. The default mode of
PHY, and the hardware configuration pins can then be used with operation is shown in Table 16. If an alternative mode of operation
their main pin function. These pins can be used in unmanaged or is required, refer to Table 17 for the suggested external pin control.
managed configuration. Table 16. Default Hardware Configuration Modes
The unmanaged configuration refers to the ADIN2111 PHY param- Hardware Configuration Pin Function Default Mode (Pin Floating)
eters being configured by the hardware configuration pins. This Port 1 Software Power-Down Mode Enabled
mode can be used when the system requires a static configuration After Reset
of the ADIN2111 port settings without the need for software control. Port 2 Software Power-Down Mode Disabled
In this context, the port to port operation on the ADIN2111 is After Reset
configured in cut through. Port 1 Transmit Amplitude 1.0 V p-p and 2.4 V p-p
The managed configuration refers to the full control of the Port 2 Transmit Amplitude 1.0 V p-p and 2.4 V p-p
ADIN2111 using software via the SPI. The two PHYs, MAC layer, SPI Protocol OPEN Alliance
and the store and forward switch can be configured in software. SPI CRC/Protection Enabled
The configuration pins can be connected to the external host or
hardware configured using pull-up/pull-down resistors. When active, Table 17. Recommended Control for Hardware Configuration Pins
the host controller can override any of the ADIN2111 hardware Required Managed Configuration Unmanaged Configuration
configurations set by the hardware pins after power-up, hardware Pin Level Options Options
reset, or software reset. High 4.7 kΩ external pull-up resistor 4.7 kΩ external pull-up resistor
UNMANAGED APPLICATIONS Host GPIO output high1
Low External pull-down resistor External pull-down resistor
In unmanaged applications, it is possible to configure the ADIN2111 Host GPIO output low1 Floating pin2
using the hardware configuration pins without any software inter- Host GPIO tristated2
vention. In that case, the ADIN2111 is configured by default in cut
Floating pin2
through mode (port to port forwarding).
1 A low value series resistor is recommended.
Software power-down after reset must be disabled on both port 2 An external pull-down resistor is recommended.
PHYs for unmanaged applications, or the ADIN2111 associated
PHY remains in power-down indefinitely, because the device can
only exit power-down from register operation using the SPI (see the Software Power-Down After Reset
Software Power-Down Mode section).
The P1_SWPD_EN (respectively P2_SWPD_EN) hardware config-
MANAGED APPLICATIONS uration pin signal is used to enable or disable the software pow-
In managed applications, the ADIN2111 can be configured by a er-down after reset feature for PHY 1 (respectively PHY 2). The
host controller via the SPI. The host controller can dynamically power-down bit (CRSM_SFT_PD) in the software power-down con-
configure the device as required by the application. trol register is set based on the Px_SWPD_EN signal status during
power-up, hardware reset, or software reset. CRSM_SFT_PD can
In managed applications, the software power-down after reset func- also be set using the SPI via the MDIO bridge to enable software
tionality can be enabled because the host controller brings the power-down after reset.
ADIN2111 to active mode using the management interface.
The P1_SWPD_EN pin has a weak internal pull-down resistor.
HARDWARE CONFIGURATION PIN Therefore, by default, the ADIN2111 Port 1 is configured with
FUNCTIONS software power-down after reset enabled.
The following functions are configurable from the ADIN2111 hard- The P2_SWPD_EN pin has a weak internal pull-up resistor. There-
ware configuration pins: fore, by default, the ADIN2111 Port 2 is configured with software
power-down after reset disabled.
► PHY 1 software power-down mode after reset
► PHY 2 software power-down mode after reset
When software power-down after reset is enabled on a port, the and the value of B10L_TX_LVL_HI_ABLE is 0. In that case, the
associated ADIN2111 PHY enters software power-down after pow- associated register on the respective PHY cannot be changed
er-up, hardware reset, or software reset. Software power-down through the SPI. That is, 2.4 V p-p operation is not possible on the
provides a lower power mode where most of the ADIN2111 internal respective PHY if it has been configured for the 1.0 V p-p level only
modules are turned off. via the Px_TX2P4_EN pin.
The ADIN2111 PHYs can be configured to exit power-down by The Px_TX2P4_EN pins have a weak internal pull-down resistor.
setting the CRSM_SFT_PD bit to 0 using the SPI via the MDIO Therefore, by default, the ADIN2111 is configured to support both
bridge. 1.0 V p-p and 2.4 V p-p voltage levels on Port 1 and Port 2.
Following a power-up, hardware reset, or software reset, and if Table 19. Port 1 Transmit Amplitude Selection (Hardware Configuration)
autonegotiation is enabled and software power-down after reset is P1_TX2P4_EN Transmit Amplitude Selection
disabled, the respective ADIN2111 PHY starts autonegotiation and 0 1.0 V p-p and 2.4 V p-p
tries to bring up a link. 1 1.0 V p-p
Table 18. Port 1/Port 2 Software Power-Down (Hardware Configuration) Table 20. Port 2 Transmit Amplitude Selection (Hardware Configuration)
Px_SWPD_EN Software Power-Down Configuration P2_TX2P4_EN Transmit Amplitude Selection
0 PHY in software power-down after reset 0 1.0 V p-p and 2.4 V p-p
1 PHY not in software power-down after reset 1 1.0 V p-p
(PHY 1) and P1_SWPD_EN (PHY 2) hardware configuration pin Each PHY of the ADIN2111 can either be configured to advertise
signals. support of both 1.0 V p-p and 2.4 V p-p transmit level operation or
to advertise support of only 1.0 V p-p transmit level operation. Refer
Switch Initialization to the Transmit Level Mode Advertisement section for more details.
After power-up, hardware reset, or software reset, the ADIN2111 1.0 V p-p transmit level operation is required for intrinsically safe
switch can be configured via the SPI. Configure the Interrupt Mask operation.
Register 0 (IMASK0) register and the mask bits for driving the
interrupt pin (IMASK1) register to enable interrupts as required. Enable High Voltage Transmit Ability
Write CONFIG0 and CONFIG2 to set up the required functionality The high voltage transmit ability is set on Port 1 (respectively Port
of the switch. For example, set the OPEN Alliance chunk size or 2) using the P1_TX2P4_EN (respectively P2_TX2P4_EN) hardware
enable cut through, if required. configuration pin signal. This signal internally sets the high voltage
transmit ability bit B10L_TX_LVL_HI_ABLE (read only), for each
When the switch is configured, write 1 to the SYNC field in the PHY, as described in the Transmit Amplitude section.
CONFIG0 register to indicate that the switch configuration is com-
plete. Enable 1.0 V p-p and 2.4 V p-p Transmit Levels
By default, the switch drops all frames after power up. The default To allow both 1.0 V p-p and 2.4 V p-p transmit level operation,
frame forwarding operation of the switch can be configured via set the 10BASE-T1L high level transmit operating mode ability bit
the P1_FWD_UNK2P2, P2_FWD_UNK2P1, P1_FWD_UNK2HOST, (AN_ADV_B10L_TX_LVL_HI_ABL) in the BASE-T1 autonegotiation
and P2_FWD_UNK2HOST bits in the Configuration Register 2 advertisement register, Bits[47:32] (AN_ADV_ABILITY_H) to 1 to
(CONFIG2). The filtering table also needs to be configured. See the indicate that the device is capable of 2.4 V p-p transmit level
Frame Forwarding on Receive section for more details. operation. A 3.3 V supply is required on the AVDD_H pins.
Configuring the Device for Linking Set 2.4 V p-p Transmit Level as Preferred
After power-up or reset, configure the ADIN2111 for the desired If 2.4 V p-p transmit level operation is preferred on PHY 1 or PHY
operation for linking. The ADIN2111 may already be configured as 2, set the 10BASE-T1L high level transmit operating mode request
required by the hardware configuration pins, but greater control is bit (AN_ADV_B10L_TX_LVL_HI_REQ) in the BASE-T1 autonego-
available using the management registers via the SPI. tiation advertisement register, Bits[47:32] (AN_ADV_ABILITY_H) to
1 on the respective PHY.
The autonegotiation process is used to match the operating mode
between a local and remote PHY. For example, autonegotiation is Set 1.0 V p-p Transmit Level as Preferred
used to ensure that the modes agree between the two devices on
which PHY operates as master and which as slave. Autonegotiation If 1.0 V p-p transmit level operation is preferred on PHY
is also used to match the transmit level between the two PHYs. 1 or PHY 2, set the AN_ADV_B10L_TX_LVL_HI_REQ bit in
the BASE-T1 autonegotiation advertisement register, Bits[47:32]
Autonegotiation is enabled by default for the ADIN2111 PHYs, (AN_ADV_B10L_TX_LVL_HI_REQ) to 0 on the respective PHY.
and it is strongly recommended to always keep autonegotiation
enabled. Autonegotiation is defined by the IEEE standard and in- Note that autonegotiation determines the transmit level at which the
cludes a number of mechanisms to ensure robust linking operation link operates.
between PHYs. Autonegotiation is the fastest way to bring up a link.
Enable 1.0 V p-p Transmit Level Only
Configuration of Transmit Level Mode If it is required to only operate the PHY 1 or PHY 2 at the 1.0
The ADIN2111 PHYs can support transmit level operation at either V p-p transmit level operation, set the 10BASE-T1L high level trans-
1.0 V p-p or 2.4 V p-p if the B10L_TX_LVL_HI_ABLE bit in mit operating mode ability bit (AN_ADV_B10L_TX_LVL_HI_ABL)
the 10BASE-T1L PMA status register (B10L_PMA_STAT) on the in the BASE-T1 autonegotiation advertisement register, Bits[47:32]
respective PHY is set to 1 and a 3.3 V supply is provided on the (AN_ADV_ABILITY_H) to 0 so that the 2.4 V p-p transmit level
AVDD_H pins. The higher transmit level can support longer reach operation is not advertised.
but also has higher power consumption. In this case, autonegotiation can only resolve to the 1.0 V p-p
The ADIN2111 PHYs can support 1.0 V p-p transmit level operation transmit level operation, irrespective of the setting that the remote
with a 1.8 V supply on the AVDD_H pins at very low power PHY advertises. For very long cable lengths, depending on the
consumption. characteristics of the cable, it may not be possible to bring up a link
at the 1.0 V p-p operation.
When a high level transmit is disabled on PHY1 (respectively See the Transmit Amplitude Advertisement section for more details.
PHY2) through the P1_TX2P4_EN (respectively P2_TX2P4_EN)
pin signal, the AVDD_H supply can be supplied from either 1.8 V or Read Link Partner Advertised Transmit Level
3.3 V for the 1.0 V p-p transmit level operation.
For each PHY, the link partner advertised transmit infor-
mation can be read from the transmit operating mode
Transmit Level Mode Advertisement ability bit (AN_LP_ADV_B10L_TX_LVL_HI_ABL) and the link
partner high level transmit operating mode request bit
Enable High Voltage Transmit Ability
(AN_LP_ADV_B10L_TX_LVL_HI_REQ) in the BASE-T1 autone-
The AVDD_H power rail must be provided with a 3.3 V supply for gotiation link partner base page ability register, Bits[47:32]
the ADIN2111 to support the 2.4 V p-p transmit level on Port 1 and (AN_LP_ADV_ABILITY_H) of the respective PHY. These bits are
Port 2. valid when autonegotiation completes (AN_COMPLETE = 1 in the
BASE-T1 autonegotiation status register).
The high voltage transmit ability is enabled on the ADIN2111 PHY
1 (respectively PHY 2) by setting the P1_TX2P4_EN (respective- See the Transmit Amplitude Advertisement section for more details.
ly P2_TX2P4_EN) hardware configuration pin signal low during
power-up, hardware reset, or software reset. The 10BASE-T1L Completion of Autonegotiation
high voltage transmit ability bit (B10L_TX_LVL_HI_ABLE) in the
10BASE-T1L PMA status register (B10L_PMA_STAT) of the re- When autonegotiation completes on PHY 1 or PHY 2, the autone-
spective PHY is set automatically to the defined hardware configu- gotiation complete indication register bit (AN_LINK_GOOD) in the
ration as follows: extra autonegotiation status register (AN_STATUS_EXTRA) is set
to 1 on the respective PHY. This bit indicates the completion of the
► B10L_TX_LVL_HI_ABLE = 0: 1.0 V p-p only ability autonegotiation sequence and that the enabled PHY link is setting
► B10L_TX_LVL_HI_ABLE = 1: 1.0 V p-p and 2.4 V p-p ability up or active.
See the Configuration of Transmit Level Mode section for more When autonegotiation completes and the link is up, the autone-
details. gotiation complete register bit (AN_COMPLETE in the BASE-T1
autonegotiation status register) is set to 1, and the contents of the
Advertise High Voltage Transmit Ability following registers are valid on the respective PHY:
To advertise the high voltage transmit ability during autonegotia- ► BASE-T1 autonegotiation advertisement registers:
tion between PHY 1 (respectively PHY 2) and a link partner, ► BASE-T1 autonegotiation advertisement register, Bits[15:0]
set the 10BASE-T1L high level transmit operating mode ability (AN_ADV_ABILITY_L)
bit (AN_ADV_B10L_TX_LVL_HI_ABL) in the BASE-T1 autonegotia- ► BASE-T1 autonegotiation advertisement register, Bits[31:16]
tion advertisement register, Bits[47:32] (AN_ADV_ABILITY_H) to 1 (AN_ADV_ABILITY_M)
on PHY 1 (respectively PHY 2). This bit can only be set if the ► BASE-T1 autonegotiation advertisement register, Bits[47:32]
ADIN2111 PHY 1 (respectively PHY 2) has the ability to transmit (AN_ADV_ABILITY_H)
in high voltage mode (B10L_TX_LVL_HI_ABLE = 1 in the 10BASE- ► BASE-T1 autonegotiation link partner base page ability registers:
T1L PMA status register).
► BASE-T1 autonegotiation link partner base page ability regis-
High voltage transmit ability only enables the ADIN2111 PHY 1 ter, Bits[15:0] (AN_LP_ADV_ABILITY_L)
(respectively PHY 2) to advertise support for both the 2.4 V p-p ► BASE-T1 autonegotiation link partner base page ability regis-
and 1.0 V p-p levels. The selected level is determined by autonego- ter, Bits[31:16] (AN_LP_ADV_ABILITY_M)
tiation with the link partner. ► BASE-T1 autonegotiation link partner base page ability regis-
See the Transmit Amplitude Advertisement section for more details. ter, Bits[47:32] (AN_LP_ADV_ABILITY_H)
a second time to determine if the link status has come up in the to the autonegotiation restart bit (AN_RESTART) in the BASE-T1
interim (see the Latch Low Registers section). autonegotiation control register (AN_CONTROL) of the respective
PHY.
If the link is dropped, the autonegotiation process restarts automati-
cally. Autonegotiation can be restarted by a request through a write
LOOPBACK MODES the SPI host, and can therefore be used to verify MAC interface
connectivity. Set the MAC_IF_LB_EN bit in the MAC interface loop-
Each of the two PHY cores on the ADIN2111 provides the following backs configuration register (MAC_IF_LOOPBACK) to 1 on PHY 1
loopback modes: (respectively PHY 2) to enable MAC interface loopback mode on
► Physical medium attachment (PMA) loopback PHY 1 (respectively PHY 2).
► Physical coding sublayer (PCS) loopback If the MAC_IF_LB_TX_SUP_EN bit in MAC_IF_LOOPBACK is set
► MAC interface loopback (enabled by default) on PHY 1 (respectively PHY 2), the transmis-
► MAC interface remote loopback sion of the signal received on the PHY MAC interface is not
► MAC loopback transferred to the ADIN2111 PHY 1 (respectively PHY 2) core.
► Host processor loopback
MAC Interface Remote Loopback
These loopback modes test and verify various functional blocks
MAC interface remote loopback mode requires a link up with
within each PHY. The use of a frame generator and frame checkers
a remote PHY and enables looping of the data received on
allows completely self contained in-circuit testing of the digital and
the ADIN2111 PHY 1 (respectively PHY 2) to the remote PHY.
analog datapaths within each PHY core.
This linking allows a remote PHY to verify a complete link
by ensuring that the PHY receives the proper data. Set the
PMA Loopback MAC_IF_REM_LB_EN bit in the MAC interface loopbacks configu-
The PHY 1 and PHY 2 can be configured in PMA loopback mode. ration register (MAC_IF_LOOPBACK) to 1 on PHY 1 (respectively
In that case, the MDI must be left open circuit, thereby transmitting PHY 2) to enable MAC interface remote loopback mode.
into an unterminated connector or cable. In this mode, the signal If the MAC_IF_REM_LB_RX_SUP_EN bit in MAC_IF_LOOPBACK
transmitted from the respective ADIN2111 PHY is echoed back from is set (set by default), the data received by the ADIN2111 PHY 1
the open 10BASE-T1L MDI. This test mode is an implementation (respectively PHY 2) from the MDI pins is not transferred to the
of the PMA local loopback function defined in Subclause 146.5.6 of ADIN2111 MAC.
the IEEE Standard 802.3cg. Remove any cable connected to the
MDI to improve the test mode accuracy. MAC Loopback
If configured in PMA loopback mode, the respective PHY must
MAC loopback mode loops the data received on the MAC transmit
be configured in forced link configuration mode (autonegotiation
channel of Port 1 or Port 2 back to the SPI host.
disabled). To enable PMA loopback mode, set the 10BASE-T1L
PMA loopback enable bit ( B10L_LB_PMA_LOC_EN ) to 1 in the MAC loopback mode can be enabled on PORT1 by setting the
10BASE-T1L PMA control register (B10L_PMA_CNTRL) of the MAC loopback bit (P1_LOOPBACK_EN) to 1 in the P1 MAC loop-
respective PHY. back enable register (P1_LOOP).
PCS Loopback MAC loopback mode can be enabled on PORT2 by setting the
MAC loopback bit (P2_LOOPBACK_EN) to 1 in the P2 MAC loop-
PCS loopback mode loops the transmit data back to the receiver back enable register (P2_LOOP).
within the PCS block at the input stage of the PHY 1 (respectively
PHY 2) digital block. Setting the B10L_LB_PCS_EN bit to 1 in the Host Processor Loopback
10BASE-T1L PCS control register (B10L_PCS_CNTRL) on PHY
1 (respectively PHY 2) enables PCS loopback mode on PHY 1 Outside of the loopback modes associated with the PHY cores
(respectively PHY 2). within the ADIN2111, the host processor can be used to create a full
MAC loopback. In a full MAC loopback, whatever frame is received
When the PCS loopback mode is enabled on PHY 1 (respectively from the MAC is transmitted back to the MAC, as shown in Figure
PHY 2), no signal is transmitted to the PHY 1 (respectively PHY 2) 10.
MDI pins.
FRAME GENERATOR AND CHECKER frame checker can be configured to check and analyze received
frames from either the MAC interface or the PHY using the frame
The two PHYs on the ADIN2111 can be configured to generate checker transmit select bit (FC_TX_SEL) in the frame checker
frames and to check received frames (see Figure 11). The generat- transmit select register (FC_TX_SEL). The frame checker reports
ing and checking functions can be used together or independently. the number of frames received, CRC errors, and various frame
If the ADIN2111 PHY 1 (respectively PHY 2) transmitted frames errors. The frame checker frame counter registers and the frame
are looped back at the remote end, the frame checker on PHY checker error counter registers count these events.
1 (respectively PH Y2) can be used to check the echoed self
generated frames. Error Counters
Frame Generator The frame checker counts the number of CRC errors, and
these errors are reported in the receive error count register
When the frame generator is enabled, the MAC interface is ignored (RX_ERR_CNT). To ensure synchronization between the frame
and the frame generator data is used for transmission on the checker error counter and frame checker frame counters, all of
MDI pins. To use the frame generator, the diagnostic clock must the counters are latched when the receive error counter register
also be enabled using the CRSM_DIAG_CLK_EN bit in the CRSM is read. Therefore, when using the frame checker, read the re-
diagnostics clock control register (CRSM_DIAG_CLK_CTRL). ceive error counter first, and then read all other frame counters
The frame generator control registers configure the type of frames and error counters. A latched copy of the receive frame counter
to be sent (for example, random data, all 1s), the frame length, and register is available in the frame checker count higher register
the number of frames to be generated. (FC_FRM_CNT_H) and the frame checker count low register
(FC_FRM_CNT_L).
The generation of the requested frames starts by enabling the
frame generator by setting the FG_EN bit in the frame generator In addition to CRC errors, the frame checker counts frame length
enable register (FG_EN). errors, frame alignment errors, symbol errors, oversized frames
errors, and undersized frame errors.
When the generation of the frames completes, the frame generator
done bit, FG_DONE, is set in the frame generator done register The frame checker also counts frames with an odd number of
(FG_DONE). nibbles in the frame, and counts packets with an odd number of
nibbles in the preamble.
Frame Checker The frame checker also counts the number of false carrier events,
The frame checker is enabled by setting the frame checker enable which is a count of the number of times the bad start of stream
bit (FC_EN) in the frame checker enable register (FC_EN). The delimiter (SSD) state is entered.
FRAME GENERATOR AND CHECKER LINK 1. The ADIN2111-1 PHY 2 frames (frame generator) are sent over
TEST the 10BASE-T1L single pair cable.
2. The ADIN2111-2 PHY 1 receives frames on the MDI pins.
Using two ADIN2111 devices, the user can configure a convenient,
3. The ADIN2111-2 PHY 1 MAC interface loops the frame back.
self contained validation setup of the PHY to PHY connection.
Figure 11 shows an overview of how each device is configured. 4. The ADIN2111-2 PHY 1 frames (looped back) are sent over the
An external cable is connected between both devices, and PHY 1 10BASE-T1L single pair cable.
is generating frames using the frame generator. PHY 2 has MAC 5. ADIN2111-1 PHY 2 receives the ADIN2111-2 PHY 1 frames
interface remote loopback mode enabled using the MAC interface (looped back) on the MDI pins.
loopbacks configuration register (MAC_IF_REM_LB_EN). 6. The ADIN2111-1 PHY 2 frame checker checks the received
frames.
See the following sequence for remote loopback used across two
PHYs for self check purposes:
ADIN2111-1 PHY 1
HOST PROCESSOR
Tx
FRAME
GENERATOR
PHY PHY
MAC DIGITAL AFE
INTERFACE
FRAME
CHECKER
Rx
MDI
SCREW
TERMINALS
SPI MAC PORT 1
PHY 2
Tx
FRAME
GENERATOR
PHY PHY
MAC DIGITAL AFE
INTERFACE
FRAME
CHECKER
Rx
MDI
SCREW
TERMINALS
PORT 2
EXTERNAL
CABLE
HOST PROCESSOR
ADIN2111-2 PHY 1
Tx
FRAME
GENERATOR
PHY PHY
MAC DIGITAL AFE
INTERFACE
FRAME
CHECKER
Rx
MDI
SCREW
TERMINALS
SPI MAC PORT 1
PHY 2
Tx
FRAME
GENERATOR
PHY PHY
MAC DIGITAL AFE
INTERFACE
FRAME
CHECKER
Rx
MDI
SCREW
TERMINALS
011
PORT 2
Figure 11. Remote Loopback Used Across Two PHYs for Self Check Purposes
Figure 13. Supplies and Capacitors for Multiple Supply 2.4 V p-p and 1.0 V
p-p Transmit Mode
Single-Supply Configuration
For single-supply operation, the same rail can be used to supply
the ADIN2111 AVDD_H, AVDD_L, and VDDIO supply rails. The
DVDDx_1P1 1.1 V rails can be derived internally or alternatively
provided by an external 1.1 V rail. This configuration is shown in
Figure 14.
Figure 12. Supplies and Capacitors for Forced 1 V p-p Transmit Mode
The output current for the LED_0 and LED_1 pins is 8 mA with
VDDIO = 3.3 V (see the Specifications section for details). For
higher current requirements, consider using the circuit described in
the Transistor Controlled LED section.
ADIN2111
Figure 14. Supplies and Capacitors for Single-Supply 2.4 V p-p Transmit RPD1
Mode R1
Px_LED_1
RPD
ETHERNET DAISY CHAIN, LINE, AND RING
NETWORK TOPOLOGIES
R0
The ADIN2111 is targeted at building automation and factory auto- Px_LED_O
RPU
mation applications that deploy 10BASE-T1L in a daisy chain, line,
RPD0
or ring network topology. These network topologies significantly VDDIO
reduce the amount of cabling required when compared to a star
016
network.
Figure 16. Active High LED Polarity Configuration
The ADIN2111 integrates two 10BASE-T1L PHY and a store and
forward switch to provide a single chip, low power solution for Active Low LED Polarity
adding devices to a daisy chain, line, or ring network topology. The
ADIN2111 provides an industry standard SPI for the device connec- In the active low configuration, the Px_LED_x pin can drive an
tion. Using SPI eases the requirements for the host processor and external LED from the cathode side. Select the R0 and R1 resistors
gives the user more choices to optimize a device for power, cost, to control the LED current (refer to the LED specifications in Table
and performance. 1 for information). External pull-up resistors (RPU0 and RPU1) with a
value of 4.7 kΩ are recommended.
SPI HOST SPI HOST SPI HOST
VDDIO
ADIN1100 ADIN2111
CONTROLLER
ADIN2111 ADIN2111 ADIN2111 RPU1
R1
Px_LED_1 VDDIO
RPD
015
10BASE-T1L LINKS
VDDIO
polarity register (LED_POLARITY). The circuits described in the
Active High LED Polarity, Active Low LED Polarity, and Autosense Figure 17. Active Low LED Polarity Configuration
Transistor Controlled LED capacitors, and ground as close to the ADIN2111 as possible. Con-
sult individual crystal vendors for recommended load information
Figure 18 displays a typical configuration where the LED current and crystal performance specifications.
required is higher than what the Px_LED_0 and Px_LED_1 pins
can supply. The circuit operates using the active high LED mode. The crystal specification defines CL. CPCB1 and CPCB2 are the
An external transistor, such as an N channel metal-oxide semicon- parasitic impedances of the PCB between XTAL_I/CLK_IN and
ductor field effect transistor (MOSFET), can be used. The transistor ground, and between XTAL_O and ground, respectively. CX1 is the
must be selected so that the gate input capacitance is not sinking capacitor connected to XTAL_I/CLK_IN, and CX2 is the capacitor
current above the maximum rating of the Px_LED_x pin during connected to XTAL_O.
the actuation. Refer to the transistor technical specifications for Assuming the following:
information. If required, the inrush current can be reduced by
placing a resistance between the transistor gate and the ADIN2111 ► CPCB1 ≈ CPCB2 ≈ CPCBx
pin, and/or adding a parallel capacitor between the GND and the ► CX1 ≈ CX2 ≈ CXx
Px_LED_x pin. The additional resistor and capacitor values must be
defined based on the transistor selection. Then, CXx = 2 × CL – CPCBx – 3 pF
Select the R0 and R1 resistors to control the LED current (refer to Choose precision capacitors for CXx with low appreciable tempera-
the selected LED and transistor specifications of the manufacturer ture coefficients to minimize frequency errors.
for information). To ensure minimum current consumption and to minimize stray ca-
External pull-down resistors (RPD0 and RPD1) with a value of 4.7 kΩ pacitances, make the connections between the crystal, capacitors,
are recommended. In Figure 18, VCC is the power supply used to and ground as close to the ADIN2111 as possible.
supply the LEDs. 25MHz
CPCB1 CPCB2
CX1 CX2
P2_CLK_IN
CLK25_REF
XTAL_I/CLK_IN
XTAL_O
ADIN2111
019
Figure 18. Transistor Controlled LED Configuration
Figure 19. Crystal Oscillator Connection
Autosense Polarity
In autosense mode, the polarity of the LED is automatically detect-
ed during power-up, hardware reset, or software reset. LED_0
(internal pull-up) and LED_1 (internal pull-down) have different
autosense behaviors due to their internal pull-up and pull-down
configurations. Use one of the configurations described in the
Active High LED Polarity, Active Low LED Polarity, and Transistor
Controlled LED sections so that the two LEDs can be controlled the
same way.
COMPONENT RECOMMENDATIONS
Crystal
The typical connection for an external crystal (XTAL) is shown
in Figure 19. To ensure minimum current consumption and to min-
imize stray capacitances, make connections between the crystal,
analog.com Rev. 0 | 33 of 125
Data Sheet ADIN2111
APPLICATIONS INFORMATION
P2_CLK_IN
CLK25_REF
XTAL_O
XTAL_I/CLK_IN
If VS p-p < 2.5 V p-p,
► C2 is not required
► C1 = 1 nF
If VS p-p ≥ 2.5 V p-p,
ADIN2111
► C2 = 10 pF
C1 = 2.5 (13 pF + CPCBx)/(VS p-p – 2.5)
020
►
The generic SPI protocol is half duplex. Therefore, it is not possible The time stamp counter can be enabled by setting TS_EN
to write frame data into the MAC_TX register and read from the (TS_CFG register) to 1 and setting FTSE (CONFIG0 register) to
MAC_RX register at the same time. Because of this, the SPI SCLK 1.
frequency must be 25 MHz to achieve full duplex transmissions on
Ethernet at 10 Mbps. Time Stamp Timer Output
MAC Frame: Transmit and Receive The ADIN2111 supports a TS_TIMER output signal. TS_TIMER can
be started at a defined time relative to the nanoseconds counter.
The 2-byte frame header shown in Table 34 is appended to all The TS_TIMER counter start time register (TS_TIMER_START) is
transmitted and received frames. This frame header always pre- used to set the start time. TS_TIMER then toggles based on the
cedes the frame data (see Figure 21). high and low times configured in high period for the TS_TIMER
register (TS_TIMER_HI) and low period for the TS_TIMER register
Time Stamp Capture (TS_TIMER_LO).
On receive, if TIME_STAMP_PRESET is asserted, an additional Use the following sequence to enable the TS_TIMER output:
4-byte or 8-byte time stamp is provided after the 2-byte header in 1. Configure the P1_LED_0 pin to the TS_TIMER function using
Table 34 and before the data frame. This time stamp can then be the LED control register (LED_CNTRL).
stored or discarded by software when reading the receive FIFO.
2. To change the default value of the TS_TIMER from 0 to 1,
On transmit, if EGRESS_CAPTURE is set other than 00, the write 1 to TS_TIMER_DEF in the timer configuration register
ADIN2111 captures the time stamp of the transmitted frame into (TS_CFG). TS_TIMER immediately toggles from 0 to 1.
the respective TTSCxH and TTSCxL registers for Port 1, and 3. Write to the TS_TIMER_HI and TS_TIMER_LO bits to set the
P2_TTSCxH and P2_TTSCxL for Port 2. required high time and low time for the TS_TIMER output.
To enable the time stamp capture signal pin, enable the P1_LED_0 4. Configure the quantization error correction register (TS_TIM-
or TEST_1 pins to be muxed as the time stamp input pin using the ER_QE_CORR) to set the required quantization error correc-
DIGIO_TSCAPT_PINMUX bits field in the Pin Mux Configuration 1 tion.
register (DIGIO_PINMUX).
5. Write a start time to the TS_TIMER_START register. When the 2-byte size field to the frame in the FIFO. Therefore, ensure that
nanoseconds counter matches this value, TS_TIMER starts to there is sufficient space for the Ethernet frame plus the 2-byte
toggle. header and 2-byte size field.
6. To stop TS_TIMER, write 1 to the TS_TIMER_STOP bit in 2. Write the size of the frame in bytes, including the 2-byte header
TS_CFG. TS_TIMER returns to the default value as set by to the MAC transmit frame size register. If the host appended a
TS_TIMER_DEF bit in TS_CFG. To start the timer again, write frame check sequence (FCS) to the frame, this is also included
to the TS_TIMER_START register. in the size.
Note that If P1_LED_0 is used as the TS_TIMER output, the 3. Write the frame data including the 2-byte frame header to the
default value is dependent on the P1_LED_0 polarity set by the transmit FIFO using the MAC transmit register. The first byte for
LED0_POLARITY bit field. transmission is written to TXD, Bits[31:24]. The full frame can
be written in a single burst or split up into multiple smaller burst
writes. The burst write data must always be in multiples of four
Transmit Frame over SPI
bytes. That is, the last word (four bytes) can contain between
The following sequence must be followed when using the generic one byte and four bytes of valid data.
SPI protocol in store and forward mode (the ADIN2111 operates in 4. When the end of frame (EOF) byte of a frame is read from the
cut through mode on the SPI by default): transmit FIFO, the bit transmit ready asserts, and an interrupt
triggers if the TX_RDY_MASK is set.
1. Verify that there is space for the frame by reading the transmit
FIFO space register. The internal MAC internally appends a
RECEIVE: 2-BYTE FRAME HEADER READ FIRST FROM THE P1_RX REGISTER
CS
3 BYTES
CD-PROT-
SDI R/W-ADDR-TA
022
Figure 22. MAC Frame: Receive
See the following definitions: The threshold at which the frame transmit starts can be
modified via the host transmit start threshold in cut through
► Priority: indicates which priority queue the frame was received bit (HOST_TX_THRESH) in the transmit threshold register
from. Not used on transmit. Set 0 in transmitted frames. (TX_THRESH). This bit a default value of 1. Therefore, by default,
► EGRESS_CAPTURE: capture an egress time stamp into the transmit starts immediately on writing to the host transmit FIFO.
host readable egress time registers, as follows:
To ensure that the frame transmission does not under run, the host
►00: no action. transmit FIFO has to be written at a rate greater than 10 Mbps. If
► 01: capture in the pair of the TTSCAL and TTSCAH registers. the frame under runs, the host transmit under run error bit asserts.
The TTSCAA bits in the STATUS0 register assert when cap-
tured. Generic SPI Errors
► 10: capture in the pair of the TTSCBL and TTSCBH registers.
The TTSCAB bits in the STATUS0 register assert when cap- Generic SPI CRC Error
tured.
If an SPI CRC error occurs on a write to a register, the register is
► 11: capture in the pair of the TTSCCL and TTSCCH registers.
not written.
The TTSCAC bits in the STATUS0 register assert when cap-
tured. If the write is to the transmit register, the transmit FIFO is missing
► TIME_STAMP_PARITY: odd parity for the appended time stamp. data and must be cleared by the host. Similarly, a read of the
Not used on transmit. Set to 0 in transmitted frames. receive register has missing data in the receive frame, and the
► TIME_STAMP_PRESENT: on receive, the first four bytes or eight FIFO must be cleared.
bytes of data contain the time stamp for the frame. Not used on If the errored transaction was a write to a configuration register, the
transmit. Set to 0 in transmitted frames. SPI host must issue the write again. If the software does not know
► Reserved: always set to 0. which configuration, reset the MAC by writing the RST_MAC_ONLY
keys to the SOFT_RST register.
Receive Frame over SPI
Generic SPI Transmit Protocol Error (TXPE)
The following procedure (which applies to Port 1 and Port 2) must
be followed to receive an Ethernet frame when using the generic TXPE asserts when the TX_FSIZE register is written, but the MAC
SPI protocol in store and forward mode (the ADIN2111 operates in still expects further writes to the transmit register related to the
store and forward mode by default): previous frame size written to the TX_FSIZE register. This error
does not occur in the normal operation and indicates an issue
1. Set the Px_RX_RDY_MASK bit to 0 to enable an interrupt when with the software driver, for example, two consecutive writes to the
a full frame is received. TX_FSIZE register without any writes to the transmit register.
2. If the Px_RX_RDY bit is asserted, read the MAC receive frame
size register to determine the size of the received frame. In response to the assertion of TXPE, the host must clear the
3. Read the frame via the MAC receive register. It is possible to transmit FIFO.
burst read the entire frame or split it up into multiple smaller OPEN ALLIANCE SPI PROTOCOL
burst reads. The first byte of the received frame is returned
in Px_RX, Bits[31:24]. The burst read transaction must be a The OPEN Alliance SPI protocol Version 1.0 can transfer data
multiple of four bytes. Some of the last four bytes are padded over the SPI using full duplex operation, achieving a 10-Mbps
with 0s if the frame is not a multiple of four bytes in size. bidirectional frame transfer with an SPI clock frequency in the
4. Read Px_RX_RDY again. If the value of the bit is 1, another region of 12 MHz to 16 MHz or greater.
frame is available to read. Repeat from Step 3. The ADIN2111 supports the following OPEN Alliance SPI capabili-
ties. See the supported capabilities register (CAPABILITY) for more
Cut Through details:
The generic SPI protocol supports cut through mode for transmit ► Transmit FCS validation
operations. ► Cut through
Before transmitting any frames, write 1 to the transmit cut through ► IEEE 1588 time stamp capture on transmit and receive
enable bits. ► Minimum supported chunk size is 8 bytes
The OPEN Alliance SPI protocol defines two types of transactions: other, allowing for the simultaneous transmission and reception of
data transactions for Ethernet frame transfers, and control transac- different length frames. The data header of the chunk in transmit
tions for register read/write operations. frames, and the data footer in receive frames, indicate which bytes
of the payload contain valid frame data. For full information on the
A chunk is the basic element of data transactions, and it is com- OPEN Alliance SPI protocol used by the ADIN2111, refer to OPEN
posed of four bytes of overhead plus the configured payload size. Alliance 10BASE-T1x MACPHY serial interface Version 1.0.
Data transactions consist of an equal number of transmit and Note that CS must be deasserted between data transactions and
receive chunks. Chunks in both transmit and receive directions control transactions, as shown in Figure 23.
may or may not contain valid frame data independent from each
MULTIPLE CHUNKS WITHIN MCU HW/FW MCU HW/FW
SPI DATA TRANSACTION TURN AROUND SPI CONTROL TRANSACTION TURN AROUND
CS
SCK
OR OR
023
Figure 23. Ethernet Data Frame Transfer Followed by Control Transfer
Data Chunks The default size of the data chunk payload is 64 bytes. This size
can be configured to 8 bytes, 16 bytes, 32 bytes, or 64 bytes via
Transmit data chunks consist of a 4-byte header followed by the the chunk payload selector bits. The data chunk size must be con-
transmit data chunk payload, as shown in Figure 24. figured before enabling data transmission or reception. Therefore,
when the data chunk size is configured, it must not be changed
HEADER TRANSMIT ETHERNET FRAME
32 BITS PAYLOAD FROM SPI HOST ON MOSI without resetting the ADIN2111.
4 BYTES TRANSMIT HEADER AND CHUNK PAYLOAD BYTES
024
Data Chunk Transactions
Figure 24. Transmit Data Chunk
Data transactions consist of 1 to N chunks on SDO and SDI. The
Receive data chunks consist of the receive data chunk payload 4-byte data header occurs at the beginning of each transmit data
followed by a 4-byte footer, as shown in Figure 25. chunk on SDO, and the 4-byte data footer occurs at the end of
each data chunk on SDI. These headers and footers contain the
RECEIVE ETHERNET FRAME
PAYLOAD TO SPI HOST ON MOSI
FOOTER
32 BITS
information needed to determine the validity and location of the
transmitted and received frames within the data chunk payload.
025
CHUNK PAYLOAD BYTES AND 4 BYTES RECEIVER HEADER
The Ethernet frames start at any 32-bit aligned word within the
Figure 25. Receive Data Chunk payloads, as shown in Figure 26 and Figure 27.
CHUNK 1 CHUNK 2 CHUNK 3 CHUNK 4
NO TX FRAME
ETH FR1 ETH FR2 CONCATENATION
(GAP IN TX)
TX FRAME ENDS
ETH FR1 ETH FR2 AT CHUNK
BOUNDARY
026
CHUNK PAYLOAD
RX FRAME ENDS
ETH FR1 ETH FR2 AT CHUNK
BOUNDARY
027
CHUNK PAYLOAD
See the following definitions: ► EBO: end byte offset. When EV is 1, this field contains the byte
offset into the transmit data chunk payload that points to the last
► SEQ: data chunk sequence. The sequence functionality is not byte of the Ethernet frame to transmit. If EV is 0, the host must
supported by the ADIN2111. This bit must be set to 0. write this field as 0.
► NORX: no receive flag. The SPI host can set this bit to indicate ► TSC: time stamp capture. Request a time stamp capture when
to the ADIN2111 that it does not process receive frame data that the frame is transmitted onto the network. See the following:
is in the current receive data chunk. For normal operation, set
► 00: no action.
NORX to 0 to indicate that it accepts and processes any receive
frame data within the current chunk. ► 01: capture in the pair of the TTSCAL and TTSCAH registers.
► VS: vendor specific bits. The TTSCAA bits in the STATUS0 register assert when cap-
tured.
► VS[1]: unused, to be set to 0 by the host.
► 10: capture in the pair of the TTSCBL and TTSCBH registers.
► VS[0]: frame destination port.
The TTSCAB bits in the STATUS0 register assert when cap-
► 0: frame destination is Port 1. tured.
► 1: frame destination is Port 2. ► 11: capture in the pair of the TTSCCL and TTSCCH registers.
► DV: data valid flag. The SPI host uses this bit to indicate if the The TTSCAC bits in the STATUS0 register assert when cap-
current chunk contains valid transmit data (DV = 1) or not. When tured.
this bit is 0, the ADIN2111 ignores the chunk payload. ► P: parity. The parity bit is calculated over the transmit data
► SV: start valid flag. When this bit is 1, the beginning of an Ether- header. The method is odd parity.
net frame is present in the current transmit data chunk payload. ► RSVD: reserved. Always set to 0.
Do not confuse the SV bit with the start of frame delimiter (SFD)
byte described in IEEE Standard 802.3.
► SWO: start word offset. When SV is 1, this field contains the
32-bit word offset into the transmit data chunk payload that
points to the start of the new Ethernet frame. If SV is 0, the host
must write this field as 0.
► EV: end valid flag. When this bit is 1, the end of an Ethernet
frame is present in the current transmit data chunk payload.
See the following definitions: ► RTSA: receive time stamp added. This bit is set when a 32-bit
or 64-bit time stamp is added to the beginning of the SPI frame.
► EXST: extended status. This bit is set when any bit in the This bit must be 0 when SV = 0.
STATUS0 or STATUS1 registers is set and not masked.
► TXC: transmit credits. This field contains the minimum number of
► HDRB: received header bad. When this bit is set, the ADIN2111 transmit data chunks of frame data that the SPI host can write in
has received a control or data header with a parity error. a single transaction without incurring a transmit buffer overflow.
► SYNC: configuration synchronized flag. This field reflects the ► P: parity. The parity bit is calculated over the receive data
state of the SYNC bit in the CONFIG0 register. When 0, this header. The method is odd parity.
bit indicates that the ADIN2111 configuration may not be as
expected by the SPI host. Following configuration, the SPI host
OPEN Alliance SPI Cut Through Mode
sets the corresponding bit in the configuration register, which is
reflected in this field. If cut through mode from or to the host is enabled, the method to
► RCA: receive chunks available. The RCA field indicates the transfer frames remains the same as when using store and forward
minimum number of additional receive data chunks of frame data mode. However, the frame receive starts when sufficient frame data
that are available for reading beyond the current one. This field is to fill a chunk is received, and the frame transmit starts when a
0 when there is no more receive frame data pending in the buffer configured transmit threshold is reached. See the transmit threshold
of the ADIN2111 to be read. register (TX_THRESH) for details.
► VS: vendor specific. The cut through mode can be enabled via the receive cut through
► VS[1]: priority of the received frame. enable bits (RXCTE) and transmit cut through enable bits (TXCTE)
► 0: frame received via the low priority queue. in the Configuration Register 0 (CONFIG0).
► 1: frame received via the high priority queue.
On receive, the MAC returns data as it becomes available. Unlike
► VS[0]: port number for received frame. in store and forward mode, there may be empty chunks (DV = 0)
► 0: frame received from Port 1. between a start of frame (SOF) chunk and an EOF chunk.
► 1: frame received from Port 2.
If the host does not read frames fast enough to keep the receive
► DV: data valid flag. The SPI host uses this bit to indicate if the FIFO empty, the frames are then buffered in the receive FIFO as if
current chunk contains valid transmit data (DV = 1) or not. When it is operating in store and forward mode. When all the frames are
this bit is 0, the SPI host ignores the chunk payload. read, the FIFO returns to operating in cut through mode.
► SV: start valid flag. When this bit is 1, the beginning of an
Ethernet frame is present in the current transmit data chunk On transmit, the host must provide frame data at a rate fast
payload. Do not confuse the SV bit with the SFD byte described enough (>10 Mbps) to ensure that the frame does not under run
in IEEE Standard 802.3. on transmit. If the MAC under runs, TXBUE in the Status Register
► SWO: start word offset. When SV is 1, this field contains the 0 (STATUS0) asserts and the MAC stops transmitting the frame in
32-bit word offset into the receive data chunk payload that points progress and appends a bad CRC to the frame.
to the start of the new Ethernet frame. When a receive time
stamp is added to the beginning of the received frame (RTSA = Cut Through Transmit Latency
1), SWO points to the most significant byte of the time stamp. If The time interval between the start of an SPI data transaction with a
SV is 0, the host must write this field as 0. transmit header SWO of 0 (frame starts immediately in the chunk),
► FD: frame drop. When set, this bit indicates that the ADIN2111 and the time TX_EN rises on the media independent interface (MII)
MAC has detected a condition for which the SPI host must drop with an SPI frequency of 16 MHz and TX_THRESH = 1 is 4 μs. The
the received Ethernet frame. This bit is only valid at the end of a PHY transmit latency is 3.2 μs. This makes a total transmit latency
received frame (EV = 1), and must be 0 at all other times. of 7.2 μs.
► EV: end valid flag. When this bit is 1, the end of an Ethernet
frame is present in the current receive data chunk payload. Cut Through Receive Latency
► EBO: end byte offset. When EV is 1, this field contains the byte The receive latency varies based on the chunk size and the SPI
offset into the receive data chunk payload that points to the last frequency. Table 37 indicates the latency for an SPI frequency of 16
byte of the received Ethernet frame. This field is 0 when EV = 0. MHz and all supported chunk sizes.
Table 37. Receive Latency for 16 MHz for All Supported Chunk Sizes
Time to Receive a Chunk
Chunk Size of Data on the Ethernet Time to Start of Frame Total Rx Latency onto Wire Total Rx Latency to End of
(Bytes) PHY Rx Latency (μs) Wire (μs) 1 Transfer over SPI (μs) 2 (μs) First Chunk Transfer (μs) 3
64 6.4 57.6 17 81 98
32 6.4 32 9 47.4 56.4
16 6.4 19.2 5 30.6 35.6
8 6.4 12.8 3 22.2 25.2
1 Enough frame data to fill a chunk must be received before a transfer starts on the SPI. The time to receive the frame preamble is also included in this chunk.
2 Assuming that the μC is not waiting for an interrupt and that it is providing back to back OPEN Alliance data transactions on the SPI. The frame transfers start in the middle
of the chunk on average.
3 Realistically, the μC cannot use the data until it receives the receive header at the end of the chunk.
Control Transactions
Table 38. Control Command Header
D31 D30 D29 D28 D27 to D24 D23 to D8 LEN7 to LEN1 P
0 HDRB WNR AID MMS ADDR [15:0] LEN P
Control transactions consist of one or more control commands. Table 39. Register Memory Maps (MMS)
These commands are used by the SPI host to read and write MMS Memory Map Description
registers within the ADIN2111, and each one is composed of a 0 Standard control and status (SPI Address 0x00 to Address 0x20)
32-bit control command header followed by register data. See Table 1 MAC (from SPI Address 0x30)
38.
See the following definitions: Control Write
► HDRB: received header bad. When set by the ADIN2111, HDRB The ADIN2111 ignores the final 32 bits of data from the SPI host at
indicates that a header was received with a parity error. The SPI the end of the control write command. The write command and data
host must always clear this bit. The ADIN2111 ignores this value. is also echoed from the ADIN2111 back to the SPI so it can identify
► WNR: write not read. If 1, data is written to registers. Otherwise, which register write failed in the case of any bus errors.
data is read. Control write commands can write either a single register or multi-
► AID: address increment disable. When clear, the address is ple registers. When multiple registers are written, the address is
automatically post incremented by one following each register automatically post incremented.
read or write.
► MMS: memory map selector. This field selects the specific regis- When a control write command is followed by another control
ter memory map to access. See Table 39. command, the new control header must immediately follow the last
► ADDR: address of the first register within the selected memory word of the echoed register write data. The SPI host must deassert
map to access. CS following the last word of the echoed register write data when
the write command is the last command of the transaction.
► LEN: length. Specifies the number of registers to read or write.
This field is interpreted as the number of registers − 1. Therefore,
a length of 0 reads or writes a single register.
CS
SCK
028
Figure 28. OPEN Alliance Control Transaction
Control Read
The MACPHY ignores all data from the SPI host following the control header for the rest of the control read command. Control read commands
can read either a single register or multiple registers. When multiple registers are read, the address is automatically post incremented,
according to the address increment disable bit in the control header.
CS
SCK
OPEN Alliance SPI Errors on SDO for the duration of the SPI transaction. Software can
then resend the corrupted control transaction after clearing the
See the following OPEN Alliance SPI errors: header error bit.
► SPI header parity error. If a parity error is detected on a transmit ► Transmit protocol error. Occurs when the ADIN2111 detects pro-
header and there is a transmit frame in the process of transfer- tocol errors in the transfer of transmit data chunks. These errors
ring over the SPI, this frame drops. If the switch is operating in are usually due to SPI host firmware issues and do not occur in
cut through mode, the frame transmit stops and a bad CRC is normal operation. The transmit protocol error bit is set when a
appended to the frame. data header received by the ADIN2111 indicates data valid (DV =
1) without a prior start of frame indication (SV = 1), in which case
The ADIN2111 returns a fixed value of 0x4000_0000 in every the data chunk is ignored. Or, when the ADIN2111 receives two
word until CS goes high. The ADIN2111 responds with DV = 1, data headers indicating a start of frame (SV = 1) without and end
EV = 1, EBO = 0, and FD = 1 in the first data footer following a of frame (EV = 1), the ADIN2111 drops the frame data from the
new assertion of CS. previous start of frame indicator and begins accepting the frame
If there is a parity error on a control transaction, the operation data from the second start of frame indicator.
does not complete. Software can determine which transaction ► Transmit buffer overflow. Occurs when attempting to write trans-
caused the error as the ADIN2111 returns a fixed 0x4000_0000 mit frame data to the ADIN2111 when there is no transmit buffer
space available as indicated by the transmit credit field (TXC) of To forward frames with this DA to the host, set the TO_HOST bit
the previous data footer. In this condition, the ADIN2111 ignores field within the ADDR_FILT_UPRn register to 1. To apply this rule,
the transmit data chunk and sets the host transmit FIFO overflow set APPLY2PORT1 or APPLY2PORT2 to 1, accordingly.
bit, and the frame data already in the buffer is dropped.
MAC addresses can be masked using the ADDR_MSK_x registers.
► Transmit buffer under run. This error can only occur in cut For example, to receive all the MAC addresses in the range
through mode. The SPI host must always send frame data to 0x8000005A64xx, write:
the ADIN2111 faster than the network to avoid this error. When
this error occurs, the host transmit FIFO under run error bit is ► 0xFFFF to ADDR_MSK_UPR0
set, and the ADIN2111 terminates the frame being transmitted ► 0XFFFFFF00 to ADDR_MSK_LWR0
in a way that invalidates the frame. Additionally, the ADIN2111
ignores any additional frame data received from the SPI host Frames that do not match any of the 16 ADDR_FILT_x regis-
until it receives an end of frame indication (EV = 1). ters are dropped by default. The user can select to forward
► Loss of framing error. This error occurs when the CS signal is packages from unknown DAs to either the other port or the
deasserted before the expected end of the data chunk or control host with the bit fields P2_FWD_UNK2P1, P1_FWD_UNK2P2,
command. The ADIN2111 and the loss of frame error is set, any P2_FWD_UNK2HOST, and P1_FWD_UNK2HOST in the MAC con-
transmit frame in progress is dropped, and any receive frame in figuration register (CONFIG2). Figure 30 shows the forwarding
progress of being sent to the SPI host is terminated. algorithm. Frames received with a bad CRC or with RX_ERR
asserted from the PHYs, as well as runt and jabber frames, are
► Receive buffer overflow. This error occurs when the SPI host
dropped and counted.
does not read frame data from the ADIN2111 fast enough. This
error can occur both in store and forward and cut through modes. X: PORT NUMBER (1 OR 2)
When this error occurs, the ADIN2111 terminates the frame NO DA FOUND?
being received from the internal PHYs. In store and forward YES
APPLYTOPORTx = THIS PORT?
mode, no portion of the frame is transferred to the SPI host. In NO
cut through mode, the ADIN2111 terminates the frame (EV = 1) YES
PX_FWD_UNK2HOST
with frame drop set (FD = 1). YES YES TO HOST?
NO
NO
► Control data protection error. The control data protection error PX_FWD_UNK2PX?
PX_FWD_UNK2Px?
(CDPE) and the loss of frame error (LOFE) bits assert when YES TO_OTHER_PORT?
YES
protection is enabled on the OPEN Alliance SPI and there is an NO NO NO YES
TO_OTHER_PORT?
error on write data received from the host. The write does not YES NO
030
FRAME FORWARDING ON RECEIVE Figure 30. Frame Forwarding Algorithm
By default, the device drops all frames received. To receive frames, RECEIVE PRIORITY QUEUES
set up the address filtering table, or change the default operation for
all receive frames. Each of the ADIN2111 PHYs has two different FIFOs on receive: a
high priority FIFO and a low priority FIFO.
Note that frames can be forwarded from one port to two destina-
tions: the other port and the host SPI port. By default, the low priority FIFO is configured to 12 kB, and
the high priority FIFO is configured to 8 kB. The sizes of these
The device can be configured to forward up to 16 different MAC FIFOs can be changed before receiving or transmitting any
addresses based on the MAC destination address (DA). frames via the P1_RX_LO_SIZE and P1_RX_HI_SIZE (Port 1)
and P2_RX_LO_SIZE and P2_RX_HI_SIZE (Port 2) fields in the
To receive frames with a particular DA, that DA has to be pro-
FIFO_SIZE register.
grammed to one of the 16 ADDR_FILT_x registers. Each register
is 32 bits wide. Therefore, for example, to program a DA of STATISTICS COUNTERS
08:00:00:5A:64:68 to ADDR_FILT_x0, write:
The ADIN2111 has fifteen 16-bit counters on each port that incre-
► 0x0800 to ADDR_FILT_UPR0 ment on frame transmit and receive. They roll over to 0 when the
► 0x005A6468 to ADDR_FILT_LWR0 maximum value of 65535 is reached and they are not cleared on
reading. Therefore, the counters must be read every four seconds
to ensure that the counters do not roll over the last value read by SRAM ECC Error
the host.
When writing a frame to one of the ADIN2111 PHY FIFOs, the size
Table 40. Port 1 and Port 2 Statistics Counters of the frame is inserted in a 16-bit word at the front of the frame and
Name1 Description written to the FIFO. A 5-bit ECC is placed alongside the size field.
Px_RX_FRM_CNT PORT Rx frame count When this location is read from the SRAM, the ECC is checked. If
Px_RX_UCAST_CNT PORT Rx unicast frame count a double bit error is detected, the RX_ECC_ERR or TX_ECC_ERR
Px_RX_MCAST_CNT PORT Rx multicast frame count bits of the STATUS1 register are set to 1. If a double bit error is
Px_RX_BCAST_CNT PORT Rx broadcast frame count detected on reading a frame header from the receive FIFO, the
Px_RX_CRC_ERR_CNT PORT Rx CRC errored frame count frame is not transmitted.
Px_RX_ALGN_ERR_CNT PORT Rx alignment error count
In response to an ECC error, a FIFO automatically clears. All
Px_RX_PHY_ERR_CNT PORT Rx PHY error count
frames in the FIFO are lost, transmission stops, and a bad CRC is
Px_RX_LS_ERR_CNT PORT Rx long and short frame error count
appended to the frame that transmitted. The next frame received is
Px_TX_FRM_CNT PORT Tx frame count
written to a FIFO.
Px_TX_UCAST_CNT PORT Tx unicast frame count
Px_TX_MCAST_CNT PORT Tx multicast frame count Transmit Overflow Error
Px_TX_BCAST_CNT PORT Tx broadcast frame count
Px_RX_DROP_FULL_CNT PORT Rx frames dropped due to FIFO full If the host attempts to write to the transmission FIFO but there is
Px_RX_DROP_FILT_CNT PORT Rx frames dropped due to filtering not enough space, the host transmission FIFO overflow bit field
Px_RX_IFG_ERR_CNT PORT Rx frames received with interframe gap asserts (TXBOE). The software checks the space available in the
(IFG) errors transmission FIFO by reading the transmission FIFO space register
(TX_SPACE) before attempting to write to it.
1 One set of counters for each port. X represents the port number.
If this error occurs, the frame is automatically dumped, and the
software has no other requirement than to clear the TXBOE status
Receive Drop FIFO Full Counter bit. Note that frames dropped because of a transmission FIFO
Before the first byte of a received frame is written into the appro- overflow are not counted in the statistics counters.
priate receive FIFO on Port 1 (respectively Port 2), the space SPI ACCESS TO THE PHY REGISTERS
in the FIFO is checked. If there is no space for at least 256
bytes, the frame is dropped and the P1_RX_DROP_FULL_CNT The ADIN2111 provides indirect access to the two PHY manage-
(respectively P2_RX_DROP_FULL_CNT) counter increments. If ment registers via the internal SPI to MDIO bridge. The eight
there is space for at least 256 bytes in the FIFO, the logic MDIOACCn registers in the SPI register map are used to access
writes the frame to the receive FIFO. If the received frame the PHY management registers. Each MDIOACCn register corre-
exceeds 256 bytes and the receive FIFO fills, the frame sponds to an MDIO transaction.
is dropped and the P1_RX_DROP_FULL_CNT (respectively The MDC default speed is 2.5 MHz. Either a 2.5 MHz or 4.166 MHz
P2_RX_DROP_FULL_CNT) counter increments. management data clock (MDC) frequency can be selected via the
Note that if a frame is destined for both the host and the other MSPEED bits in the CONFIG2 register.
port, and the frame is not forwarded in one of the destination FIFOs The MDIO master polls the TRDONE bits of the eight MDIOACCn
because it is full, the frame is not counted as a dropped frame. registers in round robin mode. When the MDIO master detects that
FRAME RECEIVE AND TRANSMIT ERRORS one of the TRDONE fields is 0, the MDIO master starts an MDIO
transaction. When the MDIO transaction completes, the TRDONE
By default, received errored frames do not generate interrupts. bits are set to 1, and the master proceeds to check the TRDONE
These frames are dropped and counted, and the software must bits of the next MDIOACCn register.
monitor the statistics counters. The static random-access memory
(SRAM) error correction code (ECC) error and transmit overflow Each of the two PHYs has a fixed MDIO PHY address, as follows:
error are reported in the error status register (ERR_STATUS). ► PHY 1 MDIO address: 0x1
These errors result in the INT pin asserting an interrupt, unless ► PHY 2 MDIO address: 0x2
an error is masked using the error status mask register (ERR_STA-
TUS_MASK). See the SRAM ECC Error section and Transmit Note that MDIO_DEVAD is always written with the device ID of
Overflow Error section. the register being accessed. MDIO_PRTAD is set to the respective
PHY MDIO address, and MDIO_ST is written to 0x0 for Clause 45
access (this applies to all of the following Clause 45 examples).
MDIO PHY Address Determination 5. Write MDIOACC4 with MDIO_DATA = the address of Register
GHJ and TRDONE = 0x0.
The ADIN2111 allows access to the two PHY registers via an SPI to
6. Write MDIOACC5 with the write data for register GHJ,
MDIO master bridge.
MDIO_OP = 0x1, and TRDONE = 0x0.
Each of the two PHYs has a fixed MDIO PHY address, as follows: 7. Write MDIOACC6 with MDIO_DATA = the address of Register
XYZ and TRDONE = 0x0.
► MDIO PHY Address 1: PHY 1
8. Write MDIOACC7 with the write data for Register XYZ,
► MDIO PHY Address 2: PHY 2 MDIO_OP = 0x1, and TRDONE = 0x0.
Clause 45 MDIO Operation Examples 9. Host polls MDIOACC7. TRDONE = 0x1 to verify that all write
data operations are complete.
Example write to PHY 1 Register XYZ:
Example burst read starting from Register XYZ:
1. Write MDIOACC0 with MDIO_DATA = the address of Regis-
ter XYZ, MDIO_DEVAD = the device ID of Register XYZ, 1. Write MDIOACC0 with MDIO_DATA = the address of the Regis-
MDIO_PRTAD = 0x1, MDIO_OP = 0x0(ADDR), MDIO_ST = ter XYZ, MDIO_OP = 0x0(ADDR), and TRDONE = 0x0
0x0, and TRDONE = 0x0. 2. Write MDIOACC1 with MDIO_OP = 0x2(INC_RD) and
2. Write MDIOACC1 with MDIO_DATA = the value to be written to TRDONE = 0x0.
Register XYZ, MDIO_OP = 0x1(WR) and TRDONE = 0x0. 3. Write MDIOACC2 with MDIO_OP = 0x2(INC_RD) and
3. Optionally, poll MDIOACC0. TRDONE = 0x1 to determine that TRDONE = 0x0.
the write address operation has completed. 4. Write MDIOACC3 with MDIO_OP = 0x2(INC_RD) and
4. Poll MDIOACC1. TRDONE = 0x1 to determine that the write TRDONE = 0x0.
data operation completed. 5. Write MDIOACC4 with MDIO_OP = 0x2(INC_RD) and
TRDONE = 0x0.
Example read of PHY 1 Register XYZ: 6. Write MDIOACC5 with MDIO_OP = 0x2(INC_RD) and
1. Write MDIOACC0 with MDIO_DATA = the address of Register TRDONE = 0x0.
XYZ, MDIO_OP = 0x0(ADDR), and TRDONE = 0x0. 7. Write MDIOACC6 with MDIO_OP = 0x2(INC_RD) and
2. Write MDIOACC1 with MDIO_OP = 0x3(RD) and TRDONE = TRDONE = 0x0.
0x0. 8. Write MDIOACC7 with MDIO_OP = 0x2(INC_RD) and
3. Poll MDIOACC1. TRDONE = 0x1 to determine that the write TRDONE = 0x0.
data operation completed. MDIOACC1. MDIO_DATA reflects 9. Poll MDIOACC7. TRDONE = 1 to verify that all read data
the content of MDIO Register XYZ. operations are complete.
10. Read MDIOACC1. MDIO_DATA, reflects the content of Register
Example write operation followed by a read to verify the write XYZ.
operation:
11. Read MDIOACC2. MDIO_DATA, reflects the content of Register
1. Write MDIOACC0 with MDIO_DATA = the address of register XYZ. ADDR + 1.
ABC and TRDONE = 0x0. 12. Read MDIOACC3. MDIO_DATA, reflects the content of Register
2. Write MDIOACC1 with MDIO_DATA = the value to be written to XYZ. ADDR + 2.
register ABC, MDIO_OP = 0x1(WR), and TRDONE = 0x0. 13. Read MDIOACC4. MDIO_DATA, reflects the content of Register
3. Write MDIOACC2 MDIO_OP = 0x3(RD) and TRDONE = 0x0. XYZ. ADDR + 3.
4. Poll MDIOACC2. TRDONE = 0x1 to verify that all operations 14. Read MDIOACC5. MDIO_DATA, reflects the content of Register
completed. MDIO_DATA reflects the content of register ABC. XYZ. ADDR + 4.
Example of four consecutive writes. It is possible to write a com- 15. Read MDIOACC6. MDIO_DATA, reflects the content of Register
mand to all eight registers before checking any. XYZ. ADDR + 5.
16. Read MDIOACC7. MDIO_DATA, reflects the content of Register
1. Write MDIOACC0 with MDIO_DATA = the address of Register XYZ. ADDR + 6.
ABC and TRDONE = 0x0.
2. Write MDIOACC1 with the write data for register ABC, Clause 22 MDIO Operations Examples
MDIO_OP = 0x1, and TRDONE = 0x0.
Example of Clause 22 write of Register XYZ:
3. Write MDIOACC2 with MDIO_DATA = the address of Register
DEF and TRDONE = 0x0. 1. Write MDIOACC0 with MDIO_DATA = write data,
4. Write MDIOACC3 with the write data for register DEF, MDIO_DEV_AD = the address of the Register XYZ,
MDIO_OP = 0x1, and TRDONE = 0x0. MDIO_PRTAD = 0x1, MDIO_OP = 0x1(WR), MDIO_ST =
0x1(Clause 22), and TRDONE = 0x0.
analog.com Rev. 0 | 47 of 125
Data Sheet ADIN2111
SWITCH SPI
2. Poll MDIOACC0. TRDONE= 0x1 to determine that the write Recommended Register Operation
data operation is complete.
Many of the PHY registers in the ADIN2111 are defined in the IEEE
Example of Clause 22 read of Register XYZ: Standard 802.3, and the exact behavior of these registers follows
1. Write MDIOACC0 with MDIO_DEV_AD = the address of the the standard. This behavior is not always obvious and is described
Register XYZ, MDIO_PRTAD = 0x1, MDIO_OP = 0x3(RD), in this section, in addition to the recommended operation and use of
MDIO_ST = 0x1(Clause 22), and TRDONE = 0x0. the registers.
2. Poll MDIOACC0. TRDONE = 0x1 to determine that the read
Latch Low Registers
operation is complete. MDIO_DATA reflects the contents of
MDIO Register XYZ. The IEEE Standard 802.3-2018 requires certain MDIO accessi-
ble registers to exhibit latch low behavior. The idea is to allow
Example of Clause 22 write and read back of Register XYZ:
software that only intermittently reads these registers to detect
1. Write MDIOACC0 with MDIO_DATA = write data, conditions that can be transitory or short lived. For example, the
MDIO_DEV_AD = the address of the Register XYZ, AN_LINK_STATUS bit is required to latch low. When the device
MDIO_PRTAD = 0x1, MDIO_OP = 0x1(WR), MDIO_ST = exits from a reset or power-down state, the latching condition is not
0x1(Clause 22), and TRDONE = 0x0. active and the value of the AN_LINK_STATUS bit reflects the cur-
2. Write MDIOACC1 with MDIO_DEV_AD = the address of the rent status of the link. However, if the link comes up and drops, the
Register XYZ, MDIO_PRTAD = 0x1, MDIO_OP = 0x3(RD), latching condition is active. In this case, the AN_LINK_STATUS bit
MDIO_ST = 0x1(Clause 22), and TRDONE = 0x0. reads as 0 even if the link has come back up again in the interim.
3. Poll MDIOACC1. TRDONE = 0x1 to determine that the read The latching condition is only cleared when the AN_LINK_STATUS
operation is complete. MDIO_DATA reflects the contents of bit is read, ensuring the software had the opportunity to observe
MDIO Register XYZ. that the link dropped.
One implication of this latch low behavior is that, if software wishes
Contents of PHY Registers to determine the current status of the link, it must perform two reads
The PHY registers provide access to control and status information of the AN_LINK_STATUS bit back to back. The first read is needed
in the management registers. to clear any active latching condition.
The registers of the PHY Clause 45 register map are made up of Another implication is that it is important that software take ac-
four device address groupings (see Table 41) based on the MDIO count of the interaction between MDIO accessible bits that share
manageable device (MMD). Within each device address space, a register address. For example, the AN_PAGE_RX bits and
IEEE standard registers are located in register addresses between AN_LINK_STATUS bits reside at the same register address. As
0x0000 and 0x7FFF, and vendor specific registers are located in a result, reading the AN_PAGE_RX bits clears any active latching
register addresses from 0x8000 to 0xFFFF. condition associated with the AN_LINK_STATUS bits.
Table 41. Clause 45 Register Groupings IEEE Duplicated Registers
Device Address MMD Name
The IEEE Standard 802.3-2018 covers a very wide range of stand-
0x01 PMA/physical medium dependent (PMD)
ards and speeds, from 10 Mbps to 40 Gbps and higher, and
0x03 PCS includes a large number of clauses. There are registers associated
0x07 Autonegotiation with many clauses, and different PHYs can include different clauses
0x1E Vendor Specific 1 and combinations of clauses. Therefore, registers for common
Clause 45 can access to up to 32 PHYs consisting of up to 32 functions like software reset, software power-down, loopback, and
MMDs through a single MDIO interface. so on, tend to be implemented in multiple clauses.
The default value of some of the registers is determined by the In the ADIN2111, the physical implementation of these registers is
value of the hardware configuration pins, which are read just after in a single location, but they can be accessed at multiple address-
the RESET pin is deasserted. In these cases, the reset value in the es. For example, the software reset bit, can be read or written in
register table is listed as pin dependent, which allows the default all the following IEEE MMD locations and vendor specific register
operation of the ADIN2111 to be configured without having to write locations:
to it over the SPI. ► PMA_SFT_RST
► B10L_PMA_SFT_RST
► PCS_SFT_RST
► B10L_PCS_SFT_RST
Configuration Register 0
Address: 0x04, Reset: 0x00000006, Name: CONFIG0
Table 47. Bit Descriptions for CONFIG0
Bits Bit Name Description Reset Access
[31:16] RESERVED Reserved. 0x0 R
15 SYNC Configuration Synchronization. The state of this bit is reflected in the Rx footer SYNC bit. This bit defaults to 0 0x0 R/W1S
upon reset. Once written to a 1 by the SPI host, writing 0 does not clear this bit.
Immediately after any reset the SYNC bit clears to 0, RESETC is set to 1, and the interrupt pin asserts.
0: the MACPHY reset and is not configured.
1: the MACPHY is configured.
14 TXFCSVE Transmit Frame Check Sequence Validation Enable. When set, the final four octets of all Ethernet frames 0x0 R/W
received are validated. CRC_APPEND must be 0 if this bit is set. That is, the MAC must not be configured to
append a CRC to each transmitted frame.
13 CSARFE CS Align Receive Frame Enable. When set, all receive Ethernet frames data starts only at the beginning of 0x0 R/W
the first receive chunk payload following CS assertion. The start word offset (SWO) is always zero. Receive
frames can begin within any receive chunk when this bit is clear. Only applies to the OA SPI protocol.
12 ZARFE Zero Align Receive Frame Enable. When set, all receive Ethernet frames data is aligned to start at the 0x0 R/W
beginning of the receive chunk payload with an SWO of zero. Receive frames can begin anywhere within the
receive chunk payload when this bit is clear. Only applies to the Open Alliance SPI protocol.
[11:10] TXCTHRESH Transmit Credit Threshold. This field configures the number of transmit credits (TXC) of free buffer that must 0x0 R/W
be available for writing before INT asserts. Only applies to the OA SPI Protocol.
00: ≥ 1 credit.
01: ≥ 4 credits.
10: ≥ 8 credits.
Status Register 0
Address: 0x08, Reset: 0x00000040, Name: STATUS0
Table 49. Bit Descriptions for STATUS0
Bits Bit Name Description Reset Access
[31:13] RESERVED Reserved. 0x0 R
12 CDPE Control Data Protection Error. When control data read/write protection is enabled, this bit indicates that the 0x0 R/W1C
MACPHY detected an error in protected control write data received from the host. When not implemented, this
bit is reserved with a read-only value of zero. This field is only used with the OPEN Alliance SPI protocol.
11 TXFCSE Transmit Frame Check Sequence Error. This bit indicates that a frame was received over SPI from the host with 0x0 R/W1C
an invalid FCS appended. The frame is still forwarded from the device as the FCS is checked as the frame is
being transmitted.
10 TTSCAC Transmit Time Stamp Capture Available C. 0x0 R/W1C
9 TTSCAB Transmit Time Stamp Capture Available B. 0x0 R/W1C
8 TTSCAA Transmit Time Stamp Capture Available A. 0x0 R/W1C
7 PHYINT PHY Interrupt for Port 1. Host software must read the MDIO Interrupt Status registers (PHY_SUB- 0x0 R
SYS_IRQ_STATUS or CRSM_IRQ_STATUS) to determine the source of the interrupt. When exiting power-on
reset or pin reset this bit is asserted, but this field is masked from asserting an interrupt by default.
6 RESETC Reset Complete. This bit is set when the MACPHY reset is complete and ready for configuration. When it is set, 0x1 R/W1C
it generates a nonmaskable interrupt assertion on INT to alert the SPI host. Additionally, setting the RESETC bit
also sets EXST = 1 in the first Rx footer, or until this bit is cleared by action of the SPI host writing a 1.
5 HDRE Header Error. When set, this bit indicates that the MACPHY detected an invalid header received from the SPI 0x0 R/W1C
host. The invalid header is due to a parity check error. This field is only used with the OPEN Alliance SPI
protocol.
4 LOFE Loss of Frame Error. When set, this bit indicates that the MACPHY detected an early deassertion of CS prior to 0x0 R/W1C
the expected end of a data chunk or command control transaction. Note that this bit field also asserts if CDPE
asserts.
3 RXBOE Receive Buffer Overflow Error. When set, this bit indicates that the receive buffer (from the network) overflowed 0x0 R/W1C
and the receive frame data was lost.
2 TXBUE Host Tx FIFO Under Run Error. This error can only assert when cut through from the host is enabled. The host 0x0 R/W1C
software ensures this bit never asserts by writing frame data to the MAC at a rate greater than 10 Mbps. If an
under run error occurs, transmit of the current packet stops.
1 TXBOE Host Tx FIFO Overflow. The host software ensures this bit never asserts by checking the space available in 0x0 R/W1C
the Tx FIFO before writing to the Tx FIFO. If using the OPEN Alliance SPI Data protocol, the space in the Tx
Status Register 1
Address: 0x09, Reset: 0x00000000, Name: STATUS1
Table 50. Bit Descriptions for STATUS1
Bits Bit Name Description Reset Access
[31:25] RESERVED Reserved. 0x0 R/W1C
24 P2_TXFCSE CRC Mismatch on a Tx Frame on Port 2. 0x0 R/W1C
23 P2_RX_IFG_ERR Rx MAC Interframe Gap Error. If the IFG is too short, the frame is dropped on receive. The threshold 0x0 R/W1C
used for measuring the IFG on receive can be set in the P2_RX_IFG register.
22 P2_TTSCAC Transmit Time Stamp Capture Available C. 0x0 R/W1C
21 P2_TTSCAB Transmit Time Stamp Capture Available B. 0x0 R/W1C
20 P2_TTSCAA Transmit Time Stamp Capture Available A. 0x0 R/W1C
19 P2_PHYINT PHY Interrupt for Port 2. Host software must read the MDIO Interrupt Status registers (PHY_SUB- 0x0 R
SYS_IRQ_STATUS or CRSM_IRQ_STATUS) on PORT2 PHY to determine the source of the interrupt.
18 P2_RX_RDY_HI Port 2 Rx Ready High Priority. Indicates that there is a high priority frame available on Port 2. See 0x0 R
P1_RX_RDY_HI. This field does not drive the interrupt pin. This field is only used with the generic SPI
protocol.
17 P2_RX_RDY Port 2 Rx FIFO Contains Data. In store and forward mode, this field indicates that there is one or 0x0 R
more frames in Port 2 Rx FIFO. In cut through mode, this field indicates that the receive threshold
(RX_THRESH) is reached or the EOF byte of a frame is received. If there are frames in both the Rx high
and low priority FIFOs, the frame from the high priority FIFO is read first by default. This default operation
can be changed using the RX_LOW_PRI_1ST field. This field is only used with the generic SPI protocol.
[16:13] RESERVED Reserved. 0x0 R
12 TX_ECC_ERR ECC Error on Reading the Frame Size from a Tx FIFO. An uncorrectable ECC error was detected on a 0x0 R/W1C
read of the size field from the Tx FIFO. The FIFO is automatically cleared and the frame associated with
the ECC error and any other frames in the Tx FIFO is lost/dropped.
11 RX_ECC_ERR ECC Error on Reading the Frame Size from an Rx FIFO. An uncorrectable ECC error was detected 0x0 R/W1C
on a reading the frame size field from an Rx FIFO. The FIFO is automatically cleared and the frame
associated with the ECC error and other frames in the Rx FIFO is not lost/dropped.
10 SPI_ERR Detected an Error on an SPI Transaction. When using the generic SPI protocol, this field indicates that a 0x0 R/W1C
CRC error was detected. This field is not used with the OPEN Alliance Protocol. See HDRE and CDPE in
STATUS0 for OPEN Alliance SPI errors.
9 RESERVED Reserved. 0x0 R/W1C
8 P1_RX_IFG_ERR Rx MAC Interframe Gap Error. If the IFG is too short, the frame is dropped on receive. The threshold 0x0 R/W1C
used for measuring the IFG on receive can be set in the P1_RX_IFG register.
[7:6] RESERVED Reserved. 0x0 R
Transmit Time Stamp Capture Register A This field contains the upper 32 bits of the captured time stamp for
(High) when the requested frame was transmitted.
Address: 0x10, Reset: 0x00000000, Name: TTSCAH
Table 54. Bit Descriptions for TTSCAH
Bits Bit Name Description Reset Access
[31:0] TTSCH_A Transmit Time Stamp A, Bits[63:32]. 0x0 R
Transmit Time Stamp Capture Register A (Low) This field contains the lower 32 bits of the captured time stamp for
when the requested frame was transmitted.
Address: 0x11, Reset: 0x00000000, Name: TTSCAL
Table 55. Bit Descriptions for TTSCAL
Bits Bit Name Description Reset Access
[31:0] TTSCL_A Transmit Time Stamp A, Bits[31:0]. 0x0 R
Transmit Time Stamp Capture Register B This field contains the upper 32 bits of the captured time stamp for
(High) when the requested frame was transmitted.
Address: 0x12, Reset: 0x00000000, Name: TTSCBH
Table 56. Bit Descriptions for TTSCBH
Bits Bit Name Description Reset Access
[31:0] TTSCH_B Transmit Time Stamp B, Bits[63:32]. 0x0 R
Transmit Time Stamp Capture Register B (Low) This field contains the lower 32 bits of the captured time stamp for
when the requested frame was transmitted.
Address: 0x13, Reset: 0x00000000, Name: TTSCBL
Table 57. Bit Descriptions for TTSCBL
Bits Bit Name Description Reset Access
[31:0] TTSCL_B Transmit Time Stamp B, Bits[31:0]. 0x0 R
Transmit Time Stamp Capture Register C This field contains the upper 32 bits of the captured time stamp for
(High) when the requested frame was transmitted.
Address: 0x14, Reset: 0x00000000, Name: TTSCCH
Table 58. Bit Descriptions for TTSCCH
Bits Bit Name Description Reset Access
[31:0] TTSCH_C Transmit Time Stamp C Bits, [63:32]. 0x0 R
Transmit Time Stamp Capture Register C (Low) This field contains the lower 32 bits of the captured time stamp for
when the requested frame was transmitted.
Address: 0x15, Reset: 0x00000000, Name: TTSCCL
Table 59. Bit Descriptions for TTSCCL
Bits Bit Name Description Reset Access
[31:0] TTSCL_C Transmit Time Stamp C, Bits[31:0]. 0x0 R
Use this register to access the PHY registers via the SPI to MDIO bridge.
Scratch Registers
Address: 0x37 to 0x3A (Increments of 1), Reset: 0x00000000, Name: SCRATCHn
When writing the ADDR_MSK_x registers, all two register locations must be written in order for a given table entry. They must be written in
order with the UPR register written first and the LWR register written last.
P1 to P2 Tx Size Register
Address: 0xBC, Reset: 0x00000000, Name: P1TOP2_TXSIZE
Number of valid half words (16 Bits) in Port 1 Tx FIFO to be transmitted to Port 2.
This counter does not increment on LES because a PHY error cannot be determined via the reduced media independent interface (RMII).
Instead, PHY errors are seen as CRC errored frames in the MAC. The PHY error counter on the PHY must be read to determine if there are
PHY errors on Port 2.
P2 to P1 Tx Size Register
Address: 0xEC, Reset: 0x00000000, Name: P2TOP1_TXSIZE
Number of valid half words (16 bits) in Port 2 Tx FIFO to be transmitted to Port 1.
Transmit Time Stamp Capture Register A This field contains the upper 32 bits of the captured time stamp for
(High) when the requested frame was transmitted.
Address: 0xF0, Reset: 0x00000000, Name: P2_TTSCAH
Table 149. Bit Descriptions for P2_TTSCAH
Bits Bit Name Description Reset Access
[31:0] P2_TTSCH_A Port 2 Transmit Time Stamp A, Bits[63:32]. 0x0 R
Transmit Time Stamp Capture Register A (Low) This field contains the lower 32 bits of the captured time stamp for
when the requested frame was transmitted.
Address: 0xF1, Reset: 0x00000000, Name: P2_TTSCAL
Transmit Time Stamp Capture Register B This field contains the upper 32 bits of the captured time stamp for
(High) when the requested frame was transmitted.
Address: 0xF2, Reset: 0x00000000, Name: P2_TTSCBH
Table 151. Bit Descriptions for P2_TTSCBH
Bits Bit Name Description Reset Access
[31:0] P2_TTSCH_B Port 2 Transmit Time Stamp B, Bits[63:32]. 0x0 R
Transmit Time Stamp Capture Register B (Low) This field contains the lower 32 bits of the captured time stamp for
when the requested frame was transmitted.
Address: 0xF3, Reset: 0x00000000, Name: P2_TTSCBL
Table 152. Bit Descriptions for P2_TTSCBL
Bits Bit Name Description Reset Access
[31:0] P2_TTSCL_B Port 2 Transmit Time Stamp B, Bits[31:0]. 0x0 R
Transmit Time Stamp Capture Register C This field contains the upper 32 bits of the captured time stamp for
(High) when the requested frame was transmitted.
Address: 0xF4, Reset: 0x00000000, Name: P2_TTSCCH
Table 153. Bit Descriptions for P2_TTSCCH
Bits Bit Name Description Reset Access
[31:0] P2_TTSCH_C Port 2 Transmit Time Stamp C, Bits[63:32]. 0x0 R
Transmit Time Stamp Capture Register C (Low) This field contains the lower 32 bits of the captured time stamp for
when the requested frame was transmitted.
Address: 0xF5, Reset: 0x00000000, Name: P2_TTSCCL
Table 154. Bit Descriptions for P2_TTSCCL
Bits Bit Name Description Reset Access
[31:0] P2_TTSCL_C Port 2 Transmit Time Stamp C, Bits[31:0]. 0x0 R
This address corresponds to the 10BASE-T1L PCS control register specified in Clause 45.2.3.68a of Standard 802.3cg.
Device Address: 0x1E; Register Address: 0x0006, Reset: 0xC000, Name: MMD1_DEVS_IN_PKG2
Vendor-specific device 1 and Vendor-specific device 2 MMDs present
This address can be used to check which interrupt requests have been triggered since the last time it was read. Each bit goes high when
the associated event occurs and then latches high until it is unlatched by reading. The bits of CRSM_IRQ_STATUS go high even when the
associated interrupts are not enabled. A reserved interrupt being triggered indicates a fatal error in the system.
Device Address: 0x1F; Register Address: 0x0006, Reset: 0xC000, Name: MMD2_DEVS_IN_PKG2
Vendor Specific 1 and Vendor Specific 2 MMDs are present.
DETAIL A
(JEDEC 95)
7.10
7.00 SQ 0.30
6.90 0.25
PIN 1
INDICATOR 0.18
AREA PIN 1
INDICATOR AREA OPTIONS
37 48 (SEE DETAIL A)
36 1
0.50
BSC 5.70
EXPOSED 5.60 SQ
PAD
5.50
24 13
0.50 0.20 MIN
TOP VIEW BOTTOM VIEW
0.40
0.30 5.50 REF
0.80
0.75 END VIEW
0.05 MAX
0.70
0.02 NOM
COPLANARITY
SEATING 0.08
PLANE 0.203 REF
10-10-2018-A
PKG-004452
EVALUATION BOARDS
©2021 Analog Devices, Inc. All rights reserved. Trademarks and Rev. 0 | 125 of 125
registered trademarks are the property of their respective owners.
One Analog Way, Wilmington, MA 01887-2356, U.S.A.
Mouser Electronics
Authorized Distributor